CR Model selection and implemented WandB for SKlearn

.gitignore

@@ -0,0 +1,6 @@
+__pycache__/
+poetry.lock
+*.env
+**/.env
+.DS_Store
+**/wandb/

data/processed/PII_Customer_Personality_Analysis/artifacts/corr_matrix_PII_Customer_Personality_Analysis_correlation_matrix.csv

@@ -0,0 +1,26 @@
+,Education,Marital_Status,Income,Kidhome,Teenhome,Recency,MntWines,MntFruits,MntMeatProducts,MntFishProducts,MntSweetProducts,MntGoldProds,NumDealsPurchases,NumWebPurchases,NumCatalogPurchases,NumStorePurchases,NumWebVisitsMonth,AcceptedCmp3,AcceptedCmp4,AcceptedCmp5,AcceptedCmp1,AcceptedCmp2,Complain,target,Age
+Education,1.0,0.00601057971479844,0.12069235890931647,-0.045390101722656215,0.12012771551954766,-0.011418202661844836,0.19788629514026926,-0.08246221470202815,0.03996093529190393,-0.11474748247768826,-0.10727879120827798,-0.09708394083472364,0.026207871865264364,0.08242517900615554,0.06904876324567108,0.06779203319202283,-0.040821622200267235,0.005822767693396697,0.0588849087724573,0.03266911640198275,-0.009741208717866857,0.02147726629373715,-0.05086296084713461,0.09080601779180715,0.1710648461886182
+Marital_Status,0.00601057971479844,1.0,0.021352847397029834,-0.021718068362697428,-0.0032431226287101722,0.011137721721606375,0.009228819772649922,0.0005591923135064867,0.02780994113296931,0.03420772518449643,0.01565110733819486,0.0008966465168111868,-0.019293729796295137,0.00011772760451422034,0.014930606044176415,0.002603366103162813,-0.026657557466784578,-0.02623359458981945,0.014274398390713436,0.010256030152237389,-0.015569020204498948,0.018908848090815854,-0.005393690720407015,-0.012641150709966707,0.058228774244458686
+Income,0.12069235890931647,0.021352847397029834,1.0,-0.42866900796918467,0.019133378179405063,-0.003969755538429371,0.5786497501367387,0.4308416809908738,0.5846333567663218,0.43887135945164113,0.44074379151936854,0.3259164464972651,-0.08310089573037695,0.38787781129179055,0.5891624419343273,0.5293621402734204,-0.5530880116530972,-0.016174440058630012,0.18440036817458086,0.3359432659885299,0.27681986364264505,0.08754477410485566,-0.02722451231447758,0.13304666375157864,0.16179142819632944
+Kidhome,-0.045390101722656215,-0.021718068362697428,-0.42866900796918467,1.0,-0.03986909491269764,0.011492148877606401,-0.49733585804824904,-0.3733961018645468,-0.4392605293302491,-0.3888842203052554,-0.3780261275829725,-0.35502942464114273,0.2169130475866851,-0.3719765493465889,-0.5045006219645888,-0.5013488096627505,0.44747694055447784,0.016066022250294954,-0.16202597120826812,-0.20530460075712562,-0.17416308346026005,-0.08186792980170485,0.040977948951332006,-0.0779087218804935,-0.23361461678438017
+Teenhome,0.12012771551954766,-0.0032431226287101722,0.019133378179405063,-0.03986909491269764,1.0,0.013837883216177588,0.003746662683131421,-0.17655763847960287,-0.26112238523373393,-0.2052418665057769,-0.1630557773562015,-0.01988723381366726,0.38624630410345656,0.16207718515997702,-0.11269219774843708,0.04973702442961182,0.13124002195157045,-0.04252154241540535,0.038375727599795045,-0.19079132962891837,-0.14485535030427124,-0.015520861256563333,0.003306980000913191,-0.15390119913873612,0.35079057250867324
+Recency,-0.011418202661844836,0.011137721721606375,-0.003969755538429371,0.011492148877606401,0.013837883216177588,1.0,0.015721019423318053,-0.005843749911872586,0.022517635114411027,0.0005509232351481579,0.02510977031031601,0.01766263773496973,0.0021154508116066205,-0.005640853762668498,0.02408140757530815,-0.0004338265887671439,-0.01856364340172459,-0.03225726632386708,0.017566258881346348,-0.0004819996751676512,-0.021061220855675345,-0.0014003820869686575,0.01363667026546777,-0.19976636929150263,0.016294899725588837
+MntWines,0.19788629514026926,0.009228819772649922,0.5786497501367387,-0.49733585804824904,0.003746662683131421,0.015721019423318053,1.0,0.3870238608948813,0.5688600028034905,0.3977210502397404,0.39032580211914597,0.3927309933121961,0.008885928846518159,0.5537859390502029,0.6347527405610132,0.6400119079346188,-0.3219779006408591,0.06146322133332154,0.37314333590551424,0.473550447360861,0.3514171077786048,0.20618492919071135,-0.03947021117770349,0.24629895700789062,0.15945109606670785
+MntFruits,-0.08246221470202815,0.0005591923135064867,0.4308416809908738,-0.3733961018645468,-0.17655763847960287,-0.005843749911872586,0.3870238608948813,1.0,0.5478221664444243,0.5934310502810902,0.5716060634768749,0.39648692442903605,-0.13451209943613016,0.3020388491308559,0.4862630707589945,0.45849103147873843,-0.4187289323674573,0.014423959637837509,0.006395603242838285,0.21287107479087422,0.1918157630946309,-0.009980152400035486,-0.005324098581870991,0.12244267882212094,0.017746520210085638
+MntMeatProducts,0.03996093529190393,0.02780994113296931,0.5846333567663218,-0.4392605293302491,-0.26112238523373393,0.022517635114411027,0.5688600028034905,0.5478221664444243,1.0,0.5735740153436658,0.5351361087117303,0.35944628071880613,-0.12130771413541634,0.30709036563638303,0.7341265978631459,0.4860055452848598,-0.5394844166382349,0.018437950940138954,0.09161819640940018,0.3768671184726933,0.31307611216860326,0.043521399455980485,-0.0237819441070392,0.23774641828315587,0.03369674544450875
+MntFishProducts,-0.11474748247768826,0.03420772518449643,0.43887135945164113,-0.3888842203052554,-0.2052418665057769,0.0005509232351481579,0.3977210502397404,0.5934310502810902,0.5735740153436658,1.0,0.5838669550256257,0.4271420401285096,-0.14324108564174431,0.29968751037596947,0.5327567837134197,0.45774504320424736,-0.44642329175818335,-0.00021899209055690604,0.016105384537408233,0.19627745116954454,0.2616081098154486,0.0023448969973614508,-0.02122023035141569,0.10814510985482437,0.04042508416794455
+MntSweetProducts,-0.10727879120827798,0.01565110733819486,0.44074379151936854,-0.3780261275829725,-0.1630557773562015,0.02510977031031601,0.39032580211914597,0.5716060634768749,0.5351361087117303,0.5838669550256257,1.0,0.35744974733671053,-0.12143192773267662,0.3339372174806106,0.49513581787229005,0.4552251635983501,-0.4223708035871346,0.0017804341899794044,0.029313011665345452,0.2592298737532875,0.24510196257446312,0.010188061683200663,-0.0226412001817123,0.1161703734475162,0.02020441495744574
+MntGoldProds,-0.09708394083472364,0.0008966465168111868,0.3259164464972651,-0.35502942464114273,-0.01988723381366726,0.01766263773496973,0.3927309933121961,0.39648692442903605,0.35944628071880613,0.4271420401285096,0.35744974733671053,1.0,0.05190482939144336,0.40706566619258716,0.44242825214836723,0.3891801722259885,-0.2476905571725933,0.1249578642120278,0.024015092911599267,0.1813973755277105,0.17013156126608653,0.05073361018385326,-0.031133459344139035,0.1403316444499485,0.06420769327026418
+NumDealsPurchases,0.026207871865264364,-0.019293729796295137,-0.08310089573037695,0.2169130475866851,0.38624630410345656,0.0021154508116066205,0.008885928846518159,-0.13451209943613016,-0.12130771413541634,-0.14324108564174431,-0.12143192773267662,0.05190482939144336,1.0,0.24144031825434095,-0.012118428034188273,0.06610659381931695,0.3460483799648631,-0.023135079994654903,0.016076520490276074,-0.1842529426068536,-0.12737389187862327,-0.03798115053624632,0.0004972466506366478,0.003451073256152576,0.05866805087264471
+NumWebPurchases,0.08242517900615554,0.00011772760451422034,0.38787781129179055,-0.3719765493465889,0.16207718515997702,-0.005640853762668498,0.5537859390502029,0.3020388491308559,0.30709036563638303,0.29968751037596947,0.3339372174806106,0.40706566619258716,0.24144031825434095,1.0,0.3868676401456998,0.5162401826934668,-0.05122626307505021,0.04295782900243499,0.1629322581651734,0.14118889093609427,0.1592916660892605,0.034828595290096596,-0.016641779042421617,0.1514312334625439,0.1530513747670212
+NumCatalogPurchases,0.06904876324567108,0.014930606044176415,0.5891624419343273,-0.5045006219645888,-0.11269219774843708,0.02408140757530815,0.6347527405610132,0.4862630707589945,0.7341265978631459,0.5327567837134197,0.49513581787229005,0.44242825214836723,-0.012118428034188273,0.3868676401456998,1.0,0.517840451115637,-0.5220037739848213,0.10434509873303167,0.14018199197446132,0.3224705753164749,0.3090257184513369,0.09991528121453742,-0.0208391906218129,0.2199136123036807,0.12176397201297959
+NumStorePurchases,0.06779203319202283,0.002603366103162813,0.5293621402734204,-0.5013488096627505,0.04973702442961182,-0.0004338265887671439,0.6400119079346188,0.45849103147873843,0.4860055452848598,0.45774504320424736,0.4552251635983501,0.3891801722259885,0.06610659381931695,0.5162401826934668,0.517840451115637,1.0,-0.4323982572659749,-0.06891258923172716,0.17802019025920685,0.212953710216086,0.1787428895679233,0.08527077669680684,-0.016940707007434427,0.03624112917284234,0.12789072181374891
+NumWebVisitsMonth,-0.040821622200267235,-0.026657557466784578,-0.5530880116530972,0.44747694055447784,0.13124002195157045,-0.01856364340172459,-0.3219779006408591,-0.4187289323674573,-0.5394844166382349,-0.44642329175818335,-0.4223708035871346,-0.2476905571725933,0.3460483799648631,-0.05122626307505021,-0.5220037739848213,-0.4323982572659749,1.0,0.06130723471258901,-0.028665889635463914,-0.2778831014680211,-0.19477318052693973,-0.0073616649308966215,0.01978500588725408,-0.002208954040941986,-0.12390393683196442
+AcceptedCmp3,0.005822767693396697,-0.02623359458981945,-0.016174440058630012,0.016066022250294954,-0.04252154241540535,-0.03225726632386708,0.06146322133332154,0.014423959637837509,0.018437950940138954,-0.00021899209055690604,0.0017804341899794044,0.1249578642120278,-0.023135079994654903,0.04295782900243499,0.10434509873303167,-0.06891258923172716,0.06130723471258901,1.0,-0.07965858237443767,0.08024761492969107,0.09568286876259005,0.07170217241720112,0.008124113453170467,0.25400486323255694,-0.06178380024839474
+AcceptedCmp4,0.0588849087724573,0.014274398390713436,0.18440036817458086,-0.16202597120826812,0.038375727599795045,0.017566258881346348,0.37314333590551424,0.006395603242838285,0.09161819640940018,0.016105384537408233,0.029313011665345452,0.024015092911599267,0.016076520490276074,0.1629322581651734,0.14018199197446132,0.17802019025920685,-0.028665889635463914,-0.07965858237443767,1.0,0.3113144997868341,0.24278177008858737,0.295049565226579,-0.027651941592759625,0.18020529304447308,0.06610852367279324
+AcceptedCmp5,0.03266911640198275,0.010256030152237389,0.3359432659885299,-0.20530460075712562,-0.19079132962891837,-0.0004819996751676512,0.473550447360861,0.21287107479087422,0.3768671184726933,0.19627745116954454,0.2592298737532875,0.1813973755277105,-0.1842529426068536,0.14118889093609427,0.3224705753164749,0.212953710216086,-0.2778831014680211,0.08024761492969107,0.3113144997868341,1.0,0.407877927952029,0.22212082088409418,-0.009576350926681113,0.32337384792407603,-0.010574840069471311
+AcceptedCmp1,-0.009741208717866857,-0.015569020204498948,0.27681986364264505,-0.17416308346026005,-0.14485535030427124,-0.021061220855675345,0.3514171077786048,0.1918157630946309,0.31307611216860326,0.2616081098154486,0.24510196257446312,0.17013156126608653,-0.12737389187862327,0.1592916660892605,0.3090257184513369,0.1787428895679233,-0.19477318052693973,0.09568286876259005,0.24278177008858737,0.407877927952029,1.0,0.17663707327744818,-0.025593647329387348,0.2973447406527777,0.009610506551183966
+AcceptedCmp2,0.02147726629373715,0.018908848090815854,0.08754477410485566,-0.08186792980170485,-0.015520861256563333,-0.0014003820869686575,0.20618492919071135,-0.009980152400035486,0.043521399455980485,0.0023448969973614508,0.010188061683200663,0.05073361018385326,-0.03798115053624632,0.034828595290096596,0.09991528121453742,0.08527077669680684,-0.0073616649308966215,0.07170217241720112,0.295049565226579,0.22212082088409418,0.17663707327744818,1.0,-0.011458504341450727,0.16929370922966702,0.006716955921569739
+Complain,-0.05086296084713461,-0.005393690720407015,-0.02722451231447758,0.040977948951332006,0.003306980000913191,0.01363667026546777,-0.03947021117770349,-0.005324098581870991,-0.0237819441070392,-0.02122023035141569,-0.0226412001817123,-0.031133459344139035,0.0004972466506366478,-0.016641779042421617,-0.0208391906218129,-0.016940707007434427,0.01978500588725408,0.008124113453170467,-0.027651941592759625,-0.009576350926681113,-0.025593647329387348,-0.011458504341450727,1.0,-0.0020292937073078535,0.030407246701888446
+target,0.09080601779180715,-0.012641150709966707,0.13304666375157864,-0.0779087218804935,-0.15390119913873612,-0.19976636929150263,0.24629895700789062,0.12244267882212094,0.23774641828315587,0.10814510985482437,0.1161703734475162,0.1403316444499485,0.003451073256152576,0.1514312334625439,0.2199136123036807,0.03624112917284234,-0.002208954040941986,0.25400486323255694,0.18020529304447308,0.32337384792407603,0.2973447406527777,0.16929370922966702,-0.0020292937073078535,1.0,-0.023692119864284135
+Age,0.1710648461886182,0.058228774244458686,0.16179142819632944,-0.23361461678438017,0.35079057250867324,0.016294899725588837,0.15945109606670785,0.017746520210085638,0.03369674544450875,0.04042508416794455,0.02020441495744574,0.06420769327026418,0.05866805087264471,0.1530513747670212,0.12176397201297959,0.12789072181374891,-0.12390393683196442,-0.06178380024839474,0.06610852367279324,-0.010574840069471311,0.009610506551183966,0.006716955921569739,0.030407246701888446,-0.023692119864284135,1.0

data/processed/PII_Customer_Personality_Analysis/data/corr_matrix_PII_Customer_Personality_Analysis_correlation_matrix.csv

@@ -0,0 +1,26 @@
+,Education,Marital_Status,Income,Kidhome,Teenhome,Recency,MntWines,MntFruits,MntMeatProducts,MntFishProducts,MntSweetProducts,MntGoldProds,NumDealsPurchases,NumWebPurchases,NumCatalogPurchases,NumStorePurchases,NumWebVisitsMonth,AcceptedCmp3,AcceptedCmp4,AcceptedCmp5,AcceptedCmp1,AcceptedCmp2,Complain,target,Age
+Education,1.0,0.00601057971479844,0.12069235890931647,-0.045390101722656215,0.12012771551954766,-0.011418202661844836,0.19788629514026926,-0.08246221470202815,0.03996093529190393,-0.11474748247768826,-0.10727879120827798,-0.09708394083472364,0.026207871865264364,0.08242517900615554,0.06904876324567108,0.06779203319202283,-0.040821622200267235,0.005822767693396697,0.0588849087724573,0.03266911640198275,-0.009741208717866857,0.02147726629373715,-0.05086296084713461,0.09080601779180715,0.1710648461886182
+Marital_Status,0.00601057971479844,1.0,0.021352847397029834,-0.021718068362697428,-0.0032431226287101722,0.011137721721606375,0.009228819772649922,0.0005591923135064867,0.02780994113296931,0.03420772518449643,0.01565110733819486,0.0008966465168111868,-0.019293729796295137,0.00011772760451422034,0.014930606044176415,0.002603366103162813,-0.026657557466784578,-0.02623359458981945,0.014274398390713436,0.010256030152237389,-0.015569020204498948,0.018908848090815854,-0.005393690720407015,-0.012641150709966707,0.058228774244458686
+Income,0.12069235890931647,0.021352847397029834,1.0,-0.42866900796918467,0.019133378179405063,-0.003969755538429371,0.5786497501367387,0.4308416809908738,0.5846333567663218,0.43887135945164113,0.44074379151936854,0.3259164464972651,-0.08310089573037695,0.38787781129179055,0.5891624419343273,0.5293621402734204,-0.5530880116530972,-0.016174440058630012,0.18440036817458086,0.3359432659885299,0.27681986364264505,0.08754477410485566,-0.02722451231447758,0.13304666375157864,0.16179142819632944
+Kidhome,-0.045390101722656215,-0.021718068362697428,-0.42866900796918467,1.0,-0.03986909491269764,0.011492148877606401,-0.49733585804824904,-0.3733961018645468,-0.4392605293302491,-0.3888842203052554,-0.3780261275829725,-0.35502942464114273,0.2169130475866851,-0.3719765493465889,-0.5045006219645888,-0.5013488096627505,0.44747694055447784,0.016066022250294954,-0.16202597120826812,-0.20530460075712562,-0.17416308346026005,-0.08186792980170485,0.040977948951332006,-0.0779087218804935,-0.23361461678438017
+Teenhome,0.12012771551954766,-0.0032431226287101722,0.019133378179405063,-0.03986909491269764,1.0,0.013837883216177588,0.003746662683131421,-0.17655763847960287,-0.26112238523373393,-0.2052418665057769,-0.1630557773562015,-0.01988723381366726,0.38624630410345656,0.16207718515997702,-0.11269219774843708,0.04973702442961182,0.13124002195157045,-0.04252154241540535,0.038375727599795045,-0.19079132962891837,-0.14485535030427124,-0.015520861256563333,0.003306980000913191,-0.15390119913873612,0.35079057250867324
+Recency,-0.011418202661844836,0.011137721721606375,-0.003969755538429371,0.011492148877606401,0.013837883216177588,1.0,0.015721019423318053,-0.005843749911872586,0.022517635114411027,0.0005509232351481579,0.02510977031031601,0.01766263773496973,0.0021154508116066205,-0.005640853762668498,0.02408140757530815,-0.0004338265887671439,-0.01856364340172459,-0.03225726632386708,0.017566258881346348,-0.0004819996751676512,-0.021061220855675345,-0.0014003820869686575,0.01363667026546777,-0.19976636929150263,0.016294899725588837
+MntWines,0.19788629514026926,0.009228819772649922,0.5786497501367387,-0.49733585804824904,0.003746662683131421,0.015721019423318053,1.0,0.3870238608948813,0.5688600028034905,0.3977210502397404,0.39032580211914597,0.3927309933121961,0.008885928846518159,0.5537859390502029,0.6347527405610132,0.6400119079346188,-0.3219779006408591,0.06146322133332154,0.37314333590551424,0.473550447360861,0.3514171077786048,0.20618492919071135,-0.03947021117770349,0.24629895700789062,0.15945109606670785
+MntFruits,-0.08246221470202815,0.0005591923135064867,0.4308416809908738,-0.3733961018645468,-0.17655763847960287,-0.005843749911872586,0.3870238608948813,1.0,0.5478221664444243,0.5934310502810902,0.5716060634768749,0.39648692442903605,-0.13451209943613016,0.3020388491308559,0.4862630707589945,0.45849103147873843,-0.4187289323674573,0.014423959637837509,0.006395603242838285,0.21287107479087422,0.1918157630946309,-0.009980152400035486,-0.005324098581870991,0.12244267882212094,0.017746520210085638
+MntMeatProducts,0.03996093529190393,0.02780994113296931,0.5846333567663218,-0.4392605293302491,-0.26112238523373393,0.022517635114411027,0.5688600028034905,0.5478221664444243,1.0,0.5735740153436658,0.5351361087117303,0.35944628071880613,-0.12130771413541634,0.30709036563638303,0.7341265978631459,0.4860055452848598,-0.5394844166382349,0.018437950940138954,0.09161819640940018,0.3768671184726933,0.31307611216860326,0.043521399455980485,-0.0237819441070392,0.23774641828315587,0.03369674544450875
+MntFishProducts,-0.11474748247768826,0.03420772518449643,0.43887135945164113,-0.3888842203052554,-0.2052418665057769,0.0005509232351481579,0.3977210502397404,0.5934310502810902,0.5735740153436658,1.0,0.5838669550256257,0.4271420401285096,-0.14324108564174431,0.29968751037596947,0.5327567837134197,0.45774504320424736,-0.44642329175818335,-0.00021899209055690604,0.016105384537408233,0.19627745116954454,0.2616081098154486,0.0023448969973614508,-0.02122023035141569,0.10814510985482437,0.04042508416794455
+MntSweetProducts,-0.10727879120827798,0.01565110733819486,0.44074379151936854,-0.3780261275829725,-0.1630557773562015,0.02510977031031601,0.39032580211914597,0.5716060634768749,0.5351361087117303,0.5838669550256257,1.0,0.35744974733671053,-0.12143192773267662,0.3339372174806106,0.49513581787229005,0.4552251635983501,-0.4223708035871346,0.0017804341899794044,0.029313011665345452,0.2592298737532875,0.24510196257446312,0.010188061683200663,-0.0226412001817123,0.1161703734475162,0.02020441495744574
+MntGoldProds,-0.09708394083472364,0.0008966465168111868,0.3259164464972651,-0.35502942464114273,-0.01988723381366726,0.01766263773496973,0.3927309933121961,0.39648692442903605,0.35944628071880613,0.4271420401285096,0.35744974733671053,1.0,0.05190482939144336,0.40706566619258716,0.44242825214836723,0.3891801722259885,-0.2476905571725933,0.1249578642120278,0.024015092911599267,0.1813973755277105,0.17013156126608653,0.05073361018385326,-0.031133459344139035,0.1403316444499485,0.06420769327026418
+NumDealsPurchases,0.026207871865264364,-0.019293729796295137,-0.08310089573037695,0.2169130475866851,0.38624630410345656,0.0021154508116066205,0.008885928846518159,-0.13451209943613016,-0.12130771413541634,-0.14324108564174431,-0.12143192773267662,0.05190482939144336,1.0,0.24144031825434095,-0.012118428034188273,0.06610659381931695,0.3460483799648631,-0.023135079994654903,0.016076520490276074,-0.1842529426068536,-0.12737389187862327,-0.03798115053624632,0.0004972466506366478,0.003451073256152576,0.05866805087264471
+NumWebPurchases,0.08242517900615554,0.00011772760451422034,0.38787781129179055,-0.3719765493465889,0.16207718515997702,-0.005640853762668498,0.5537859390502029,0.3020388491308559,0.30709036563638303,0.29968751037596947,0.3339372174806106,0.40706566619258716,0.24144031825434095,1.0,0.3868676401456998,0.5162401826934668,-0.05122626307505021,0.04295782900243499,0.1629322581651734,0.14118889093609427,0.1592916660892605,0.034828595290096596,-0.016641779042421617,0.1514312334625439,0.1530513747670212
+NumCatalogPurchases,0.06904876324567108,0.014930606044176415,0.5891624419343273,-0.5045006219645888,-0.11269219774843708,0.02408140757530815,0.6347527405610132,0.4862630707589945,0.7341265978631459,0.5327567837134197,0.49513581787229005,0.44242825214836723,-0.012118428034188273,0.3868676401456998,1.0,0.517840451115637,-0.5220037739848213,0.10434509873303167,0.14018199197446132,0.3224705753164749,0.3090257184513369,0.09991528121453742,-0.0208391906218129,0.2199136123036807,0.12176397201297959
+NumStorePurchases,0.06779203319202283,0.002603366103162813,0.5293621402734204,-0.5013488096627505,0.04973702442961182,-0.0004338265887671439,0.6400119079346188,0.45849103147873843,0.4860055452848598,0.45774504320424736,0.4552251635983501,0.3891801722259885,0.06610659381931695,0.5162401826934668,0.517840451115637,1.0,-0.4323982572659749,-0.06891258923172716,0.17802019025920685,0.212953710216086,0.1787428895679233,0.08527077669680684,-0.016940707007434427,0.03624112917284234,0.12789072181374891
+NumWebVisitsMonth,-0.040821622200267235,-0.026657557466784578,-0.5530880116530972,0.44747694055447784,0.13124002195157045,-0.01856364340172459,-0.3219779006408591,-0.4187289323674573,-0.5394844166382349,-0.44642329175818335,-0.4223708035871346,-0.2476905571725933,0.3460483799648631,-0.05122626307505021,-0.5220037739848213,-0.4323982572659749,1.0,0.06130723471258901,-0.028665889635463914,-0.2778831014680211,-0.19477318052693973,-0.0073616649308966215,0.01978500588725408,-0.002208954040941986,-0.12390393683196442
+AcceptedCmp3,0.005822767693396697,-0.02623359458981945,-0.016174440058630012,0.016066022250294954,-0.04252154241540535,-0.03225726632386708,0.06146322133332154,0.014423959637837509,0.018437950940138954,-0.00021899209055690604,0.0017804341899794044,0.1249578642120278,-0.023135079994654903,0.04295782900243499,0.10434509873303167,-0.06891258923172716,0.06130723471258901,1.0,-0.07965858237443767,0.08024761492969107,0.09568286876259005,0.07170217241720112,0.008124113453170467,0.25400486323255694,-0.06178380024839474
+AcceptedCmp4,0.0588849087724573,0.014274398390713436,0.18440036817458086,-0.16202597120826812,0.038375727599795045,0.017566258881346348,0.37314333590551424,0.006395603242838285,0.09161819640940018,0.016105384537408233,0.029313011665345452,0.024015092911599267,0.016076520490276074,0.1629322581651734,0.14018199197446132,0.17802019025920685,-0.028665889635463914,-0.07965858237443767,1.0,0.3113144997868341,0.24278177008858737,0.295049565226579,-0.027651941592759625,0.18020529304447308,0.06610852367279324
+AcceptedCmp5,0.03266911640198275,0.010256030152237389,0.3359432659885299,-0.20530460075712562,-0.19079132962891837,-0.0004819996751676512,0.473550447360861,0.21287107479087422,0.3768671184726933,0.19627745116954454,0.2592298737532875,0.1813973755277105,-0.1842529426068536,0.14118889093609427,0.3224705753164749,0.212953710216086,-0.2778831014680211,0.08024761492969107,0.3113144997868341,1.0,0.407877927952029,0.22212082088409418,-0.009576350926681113,0.32337384792407603,-0.010574840069471311
+AcceptedCmp1,-0.009741208717866857,-0.015569020204498948,0.27681986364264505,-0.17416308346026005,-0.14485535030427124,-0.021061220855675345,0.3514171077786048,0.1918157630946309,0.31307611216860326,0.2616081098154486,0.24510196257446312,0.17013156126608653,-0.12737389187862327,0.1592916660892605,0.3090257184513369,0.1787428895679233,-0.19477318052693973,0.09568286876259005,0.24278177008858737,0.407877927952029,1.0,0.17663707327744818,-0.025593647329387348,0.2973447406527777,0.009610506551183966
+AcceptedCmp2,0.02147726629373715,0.018908848090815854,0.08754477410485566,-0.08186792980170485,-0.015520861256563333,-0.0014003820869686575,0.20618492919071135,-0.009980152400035486,0.043521399455980485,0.0023448969973614508,0.010188061683200663,0.05073361018385326,-0.03798115053624632,0.034828595290096596,0.09991528121453742,0.08527077669680684,-0.0073616649308966215,0.07170217241720112,0.295049565226579,0.22212082088409418,0.17663707327744818,1.0,-0.011458504341450727,0.16929370922966702,0.006716955921569739
+Complain,-0.05086296084713461,-0.005393690720407015,-0.02722451231447758,0.040977948951332006,0.003306980000913191,0.01363667026546777,-0.03947021117770349,-0.005324098581870991,-0.0237819441070392,-0.02122023035141569,-0.0226412001817123,-0.031133459344139035,0.0004972466506366478,-0.016641779042421617,-0.0208391906218129,-0.016940707007434427,0.01978500588725408,0.008124113453170467,-0.027651941592759625,-0.009576350926681113,-0.025593647329387348,-0.011458504341450727,1.0,-0.0020292937073078535,0.030407246701888446
+target,0.09080601779180715,-0.012641150709966707,0.13304666375157864,-0.0779087218804935,-0.15390119913873612,-0.19976636929150263,0.24629895700789062,0.12244267882212094,0.23774641828315587,0.10814510985482437,0.1161703734475162,0.1403316444499485,0.003451073256152576,0.1514312334625439,0.2199136123036807,0.03624112917284234,-0.002208954040941986,0.25400486323255694,0.18020529304447308,0.32337384792407603,0.2973447406527777,0.16929370922966702,-0.0020292937073078535,1.0,-0.023692119864284135
+Age,0.1710648461886182,0.058228774244458686,0.16179142819632944,-0.23361461678438017,0.35079057250867324,0.016294899725588837,0.15945109606670785,0.017746520210085638,0.03369674544450875,0.04042508416794455,0.02020441495744574,0.06420769327026418,0.05866805087264471,0.1530513747670212,0.12176397201297959,0.12789072181374891,-0.12390393683196442,-0.06178380024839474,0.06610852367279324,-0.010574840069471311,0.009610506551183966,0.006716955921569739,0.030407246701888446,-0.023692119864284135,1.0

data/processed/clustering_ml.py

@@ -0,0 +1,102 @@
+import os
+
+import pandas as pd
+import torch
+from dotenv import load_dotenv
+from sklearn.cluster import AgglomerativeClustering, KMeans
+from sklearn.metrics import silhouette_score
+from sklearn.model_selection import train_test_split
+from sklearn.pipeline import Pipeline
+from sklearn.preprocessing import StandardScaler
+
+import wandb
+
+load_dotenv()
+
+
+
+device = "mps" if torch.backends.mps.is_available() else "cpu"
+# Initialize a W&B run
+wandb.login(key=os.getenv("WANDB_API_KEY"))
+wandb.init(project="customer_personality_analysis", entity="orionai", name="clustering", job_type="train")
+
+# Load the processed data
+df_processed = pd.read_csv('df_processed.csv')
+
+# Train-Test Split
+X = df_processed.drop(columns=['target'])
+y = df_processed['target']
+
+X_train, X_test, y_train, y_test = train_test_split(X, y, test_size=0.3, random_state=42)
+
+# Log dataset artifact to W&B
+dataset_artifact = wandb.Artifact('processed_data', dataset_artifact='dataset')
+dataset_artifact.add_file('df_processed.csv')
+wandb.log_artifact(dataset_artifact)
+
+# KMeans Clustering
+kmeans = KMeans(n_clusters=2, random_state=42)
+kmeans.fit(X_train)
+kmeans_labels = kmeans.predict(X_test)
+kmeans_silhouette = silhouette_score(X_test, kmeans_labels)
+
+# Log KMeans results
+wandb.log({"KMeans Silhouette Score": kmeans_silhouette})
+
+# Agglomerative Clustering
+agg_clustering = AgglomerativeClustering(n_clusters=2)
+agg_labels = agg_clustering.fit_predict(X_test)
+agg_silhouette = silhouette_score(X_test, agg_labels)
+
+# Log Agglomerative Clustering results
+wandb.log({"Agglomerative Clustering Silhouette Score": agg_silhouette})
+
+# Summary of results
+print("\nSummary of Clustering Results:")
+print(f"KMeans Silhouette Score: {kmeans_silhouette:.2f}")
+print(f"Agglomerative Clustering Silhouette Score: {agg_silhouette:.2f}")
+
+# Log final model metrics and artifacts
+wandb.log({
+    "kmeans_silhouette_score": kmeans_silhouette,
+    "agg_silhouette_score": agg_silhouette
+})
+
+# Optional: Save the trained models and log them as artifacts
+k_model_artifact = wandb.Artifact('kmeans_model', type='model')
+wandb.log_artifact(k_model_artifact)
+
+agg_model_artifact = wandb.Artifact('agg_clustering_model', type='model')
+wandb.log_artifact(agg_model_artifact)
+
+# Finish the W&B run
+wandb.finish()
+
+# Optional: If you want to run hyperparameter sweeps with W&B
+sweep_config = {
+    'method': 'random',
+    'parameters': {
+        'n_clusters': {
+            'values': [2, 3, 4, 5]
+        }
+    }
+}
+
+sweep_id = wandb.sweep(sweep_config, project="customer-response-prediction")
+
+def sweep_train():
+    # Initialize a W&B run
+    with wandb.init() as run:
+        config = wandb.config
+        
+        # Run KMeans with the sweep's number of clusters
+        kmeans_sweep = KMeans(n_clusters=config.n_clusters, random_state=42)
+        kmeans_sweep.fit(X_train)
+        kmeans_sweep_labels = kmeans.predict(X_test)
+        kmeans_sweep_silhouette = silhouette_score(X_test, kmeans_sweep_labels)
+        
+        # Log the results
+        wandb.log({"KMeans Silhouette Score": kmeans_sweep_silhouette})
+
+# Run the sweep
+wandb.agent(sweep_id, sweep_train)

data/processed/eda.py.bak

@@ -0,0 +1,104 @@
+import pandas as pd
+import numpy as np
+import seaborn as sns
+import matplotlib.pyplot as plt
+from sklearn.preprocessing import LabelEncoder, StandardScaler
+from sklearn.feature_selection import SelectKBest, f_classif
+from sklearn.model_selection import train_test_split
+from sklearn.cluster import KMeans, AgglomerativeClustering
+from sklearn.pipeline import Pipeline
+from sklearn.metrics import silhouette_score
+
+# Load the data
+file_path = 'customer_profile_marketing.csv'
+df = pd.read_csv(file_path)
+
+# Initial Analysis
+print("Initial DataFrame Info:")
+df.info()
+print("\nInitial DataFrame Head:")
+print(df.head())
+
+# Drop irrelevant columns
+df = df.drop(columns=['Unnamed: 0', 'ID', 'Dt_Customer', 'Z_CostContact', 'Z_Revenue'])
+
+# Handle missing values (if any)
+print("Missing Values: {}", df.isna().sum())
+df = df.dropna()
+
+# Feature Engineering
+# Calculate age from 'Year_Birth'
+df['Age'] = 2024 - df['Year_Birth']  # Assuming the current year is 2024
+df = df.drop(columns=['Year_Birth'])
+
+# Convert categorical variables to numerical format using Label Encoding
+label_encoder = LabelEncoder()
+df['Education'] = label_encoder.fit_transform(df['Education'])
+df['Marital_Status'] = label_encoder.fit_transform(df['Marital_Status'])
+
+df.to_csv("./data_pre_processed_v1")
+# Correlation Analysis
+corr_matrix = df.corr()
+plt.figure(figsize=(16, 12))
+sns.heatmap(corr_matrix, annot=True, fmt='.2f', cmap='coolwarm', square=True)
+plt.title('Correlation Matrix')
+plt.show()
+
+# Feature Selection using SelectKBest
+X = df.drop(columns=['Response'])  # Features
+y = df['Response']  # Target variable
+
+# Select the top 10 features
+selector = SelectKBest(score_func=f_classif, k=10)
+X_new = selector.fit_transform(X, y)
+
+# Get the columns selected
+selected_features = X.columns[selector.get_support()]
+print("Selected Features:")
+print(selected_features)
+
+# Drop unimportant features
+df_processed = df[selected_features]
+df_processed['Response'] = y
+
+# Normalize the relevant numerical features
+scaler = StandardScaler()
+numerical_cols = df_processed.select_dtypes(include=[np.number]).columns.tolist()
+numerical_cols.remove('Response')  # Remove the target variable from the list
+
+df_processed[numerical_cols] = scaler.fit_transform(df_processed[numerical_cols])
+
+# Encoding categorical variables (already encoded with LabelEncoder)
+# No additional encoding is necessary if the categorical columns have been encoded
+
+# Train-Test Split
+X_train, X_test, y_train, y_test = train_test_split(df_processed.drop(columns=['Response']),
+                                                    df_processed['Response'],
+                                                    test_size=0.3,
+                                                    random_state=42)
+
+# Clustering Algorithms
+# 1. KMeans
+kmeans = KMeans(n_clusters=2, random_state=42)
+kmeans.fit(X_train)
+kmeans_labels = kmeans.predict(X_test)
+kmeans_silhouette = silhouette_score(X_test, kmeans_labels)
+print(f"KMeans Silhouette Score: {kmeans_silhouette:.2f}")
+
+# 2. Agglomerative Clustering
+agg_clustering = AgglomerativeClustering(n_clusters=2)
+agg_labels = agg_clustering.fit_predict(X_test)
+agg_silhouette = silhouette_score(X_test, agg_labels)
+print(f"Agglomerative Clustering Silhouette Score: {agg_silhouette:.2f}")
+
+# Summary of Results
+print("\nSummary of Clustering Results:")
+print(f"KMeans Silhouette Score: {kmeans_silhouette:.2f}")
+print(f"Agglomerative Clustering Silhouette Score: {agg_silhouette:.2f}")
+
+# Optional: If you want to use a pipeline for scaling and clustering together
+# pipeline = Pipeline([('scaler', StandardScaler()), ('kmeans', KMeans(n_clusters=2, random_state=42))])
+# pipeline.fit(X_train)
+# pipeline_labels = pipeline.predict(X_test)
+# pipeline_silhouette = silhouette_score(X_test, pipeline_labels)
+# print(f"Pipeline KMeans Silhouette Score: {pipeline_silhouette:.2f}")

data/processed/eda_code_final_fixed.py

@@ -0,0 +1,128 @@
+import os
+from datetime import datetime
+
+import matplotlib.pyplot as plt
+import numpy as np
+import pandas as pd
+import seaborn as sns
+from dotenv import load_dotenv
+from sklearn.feature_selection import SelectKBest, f_classif
+from sklearn.preprocessing import LabelEncoder, StandardScaler
+
+import wandb
+from constants import project, version
+
+load_dotenv()
+
+project = project
+version = version
+
+# Log in to W&B
+wandb.login(key=os.getenv("WANDB_API_KEY"))
+
+# Initialize W&B run
+wandb.init(project=project, entity="orionai", name="PII_customer_relationship", job_type="dataset")
+
+if os.path.exists(f'./{project}/') is False:
+    os.makedirs(f'./{project}/')
+    os.makedirs(f'./{project}/artifacts/')
+    os.makedirs(f'./{project}/data/')
+    
+print("Loading data...")
+# Load the data
+file_path = 'customer_profile_marketing.csv'
+df = pd.read_csv(file_path)
+df.rename(columns={'Response': 'target'}, inplace=True)
+
+# Log the raw data as a W&B artifact
+raw_data_artifact = wandb.Artifact('customer_profile_marketing_raw', type='dataset')
+raw_data_artifact.add_file(file_path)
+wandb.log_artifact(raw_data_artifact)
+
+# Drop irrelevant columns
+df = df.drop(columns=['Unnamed: 0', 'ID', 'Dt_Customer', 'Z_CostContact', 'Z_Revenue'])
+
+print("Cleaning Data...")
+# Handle missing values (if any)
+df_copy = df.copy()
+df = df.dropna()
+print("Dropped missing values...")
+
+# Log a table with the cleaned data (before feature engineering)
+wandb.log({"cleaned_data": wandb.Table(dataframe=df)})
+
+# Feature Engineering
+df['Age'] = 2024 - df['Year_Birth']  # Assuming the current year is 2024
+df = df.drop(columns=['Year_Birth'])
+
+print("Converting categorical variables...")
+# Convert categorical variables to numerical format using LabelEncoder
+label_encoder = LabelEncoder()
+df['Education'] = label_encoder.fit_transform(df['Education'])
+df['Marital_Status'] = label_encoder.fit_transform(df['Marital_Status'])
+# Log a table with the data after feature engineering
+wandb.log({"feature_engineered_data": wandb.Table(dataframe=df)})
+
+
+# Splitting data (train_test_split in model training files)
+df_features = df.copy()
+y = df_features['target']  # Target variable
+X = df_features.drop(columns=['target']) # Features
+
+print("Normalising numerical values...")
+
+# Normalize the relevant numerical features
+scaler = StandardScaler()
+if y.values.any():
+    numerical_cols = X.select_dtypes(include=[np.number]).columns.tolist()
+
+df_transform = scaler.fit_transform(X[numerical_cols])
+df_transform = pd.DataFrame(df_transform)
+
+# Log the processed data table
+pc_artifact = wandb.Artifact("processed_data", type="dataset")
+wandb.log_artifact(pc_artifact)
+wandb.log({"processed_data": wandb.Table(dataframe=df_transform)})
+
+print("beginning encoding...")
+pre_encoding_path = f"./{project}/data/df_pre_encoding_{project}.csv"
+# Log the processed data as a W&B artifact
+pre_encoding = df_transform.to_csv(pre_encoding_path, index=False)
+processed_data_artifact = wandb.Artifact("pre-encoding-file", type='dataset')
+processed_data_artifact.add_file(pre_encoding_path)
+wandb.log_artifact(processed_data_artifact)
+
+# Correlation Analysis
+corr_matrix = df.corr()
+plt.figure(figsize=(16, 12))
+sns.heatmap(corr_matrix, annot=True, fmt='.2f', cmap='coolwarm', square=True)
+plt.title('Correlation Matrix')
+#plt.show()
+corr_matrix_str = corr_matrix.to_string()
+print(corr_matrix_str)
+corr_matrix_plot_path = f"./{project}/artifacts/"
+corr_matrix_csv = corr_matrix.to_csv(f"./{project}/artifacts/corr_matrix_{project}_correlation_matrix.csv", index=True)
+wandb.log({"corr_matrix": corr_matrix})
+
+
+# Step 3: Log the Correlation Matrix as an Artifact in W&B
+art_path=f"./{project}/artifacts/corr_matrix_{project}_"
+artifact = wandb.Artifact(name='correlation_matrix', type='dataset', description="Correlation matrix of the dataset")
+artifact.add_file(f"./{project}/artifacts/corr_matrix_{project}_correlation_matrix.csv")
+wandb.log_artifact(artifact)
+
+# Optional: Visualize the Correlation Matrix
+plt.figure(figsize=(16, 12))
+sns.heatmap(corr_matrix, annot=True, fmt='.2f', cmap='coolwarm', square=True)
+plt.title('Correlation Matrix After Encoding')
+plt.show()
+
+print("Finishing WandB...")
+if wandb.run is not None:
+    final_csv = df_transform.to_csv(f"{project}/data/{datetime.now().strftime('%Y_%m_%d')}_{project}_v{version}.csv", index=True)
+    final_data_artifact = wandb.Artifact("final_data", type='dataset')
+    final_data_artifact.add_file(final_csv)
+    wandb.log_artifact(final_data_artifact)
+    wandb.finish()
+    print("ALL DONE...")
+

data/processed/supervized_ml.py

@@ -0,0 +1,195 @@
+import json
+import os
+import random
+from typing import Any, Dict, Tuple
+
+import pandas as pd
+import torch
+from dotenv import load_dotenv
+from sklearn.ensemble import RandomForestClassifier
+from sklearn.feature_selection import SelectKBest, f_classif
+from sklearn.linear_model import LogisticRegression
+from sklearn.metrics import (accuracy_score, classification_report,
+                             confusion_matrix, f1_score, roc_auc_score)
+from sklearn.model_selection import train_test_split
+from sklearn.preprocessing import StandardScaler
+from sklearn.svm import SVC
+
+import wandb
+
+# Load environment variables
+load_dotenv()
+
+# Load project details
+from eda_code_final_fixed import project, version
+
+
+def initialize_project(project: str, version: str) -> Tuple[pd.DataFrame, pd.Series, pd.DataFrame, pd.Series, Dict[str, Any]]:
+    """
+    Initializes a project and performs a train-test split on the processed data.
+
+    Parameters:
+        project (str): The name of the project.
+        version (str): The version of the project.
+
+    Returns:
+        tuple: A tuple containing the following:
+            - X_train (pd.DataFrame): The training features.
+            - X_test (pd.DataFrame): The testing features.
+            - y_train (pd.Series): The training targets.
+            - y_test (pd.Series): The testing targets.
+            - models (dict): A dictionary of model instances.
+    """
+    data_path = "/Users/nullzero/Documents/repos/github.com/privacy-identity/vda-simulation-medical/vda-sim-medical/data/processed/PII_Customer_Personality_Analysis/data/2024_08_25_PII_Customer_Personality_Analysis_v0.1.csv"
+    
+    # Load the processed data
+    df_processed = pd.read_csv(data_path)
+    
+    # Train-Test Split
+    X = df_processed.drop(columns=['target'])
+    y = df_processed['target']
+
+    # Select the top 10 features
+    selector = SelectKBest(score_func=f_classif, k=10)
+    X_new = selector.fit_transform(X, y)
+    selected_features = X.columns[selector.get_support()]
+    X = pd.DataFrame(X_new, columns=selected_features)
+
+    # Log the selected features to W&B
+    wandb.init(project=project, entity="orionai", name="supervized_binary_classification", job_type="supervized_train")
+    wandb.log({"selected_features": selected_features.tolist()})
+
+    # Normalize the data
+    scaler = StandardScaler()
+    X_train, X_test, y_train, y_test = train_test_split(X, y, test_size=0.1, random_state=42)
+    X_train = scaler.fit_transform(X_train)
+    X_test = scaler.transform(X_test)
+
+    # Define models
+    models = {
+        "Logistic Regression": LogisticRegression(random_state=42, max_iter=1000),
+        "Random Forest": RandomForestClassifier(random_state=42, n_estimators=100),
+        "SVM": SVC(random_state=42, probability=True)
+    }
+    
+    return X_train, X_test, y_train, y_test, models
+
+
+def training_clf(X_train: pd.DataFrame, X_test: pd.DataFrame, y_train: pd.Series, y_test: pd.Series, models: Dict[str, Any], project: str, version: str) -> Dict[str, Any]:
+    """
+    Trains and logs multiple classification models using Weights & Biases (W&B).
+
+    Args:
+        X_train (pd.DataFrame): The training features.
+        X_test (pd.DataFrame): The testing features.
+        y_train (pd.Series): The training targets.
+        y_test (pd.Series): The testing targets.
+        models (dict): A dictionary of classification models to train and log.
+        project (str): The W&B project name.
+        version (str): The model version.
+
+    Returns:
+        dict: A dictionary containing the model name, classification report, confusion matrix, accuracy, ROC AUC, and F1 score for each model.
+    """
+    results = {}
+
+    for model_name, model in models.items():
+        # Initialize a new W&B run for each model
+        run = wandb.init(project=project, entity="orionai", job_type="supervized_train", name=model_name)
+        
+        # Train the model
+        model.fit(X_train, y_train)
+        
+        # Predict and evaluate
+        y_pred = model.predict(X_test)
+        y_prob = model.predict_proba(X_test)[:, 1] if hasattr(model, "predict_proba") else None
+        accuracy = accuracy_score(y_test, y_pred)
+        roc_auc = roc_auc_score(y_test, y_prob) if y_prob is not None else None
+        f1_metric = f1_score(y_test, y_pred)
+        
+        # Log metrics
+        wandb.log({
+            "accuracy": accuracy,
+            "roc_auc": roc_auc,
+            "f1_score": f1_metric
+        })
+        
+        # Log model
+        wandb.sklearn.plot_classifier(model, X_train, X_test, y_train, y_test, y_pred, y_prob, labels=["Not Buy", "Buy"])
+        
+        # Save the model to a file
+        model_filename = f"{model_name.replace(' ', '_').lower()}_model_v{version}.pkl"
+        torch.save(model, model_filename)
+        
+        # Create and log the W&B artifact for the model
+        model_artifact = wandb.Artifact(name=f"{model_name.replace(' ', '_').lower()}_v{version}", type='model')
+        model_artifact.add_file(model_filename)
+        wandb.log_artifact(model_artifact)
+        
+        # Log classification report and confusion matrix
+        class_report = classification_report(y_test, y_pred, output_dict=True)
+        conf_matrix = confusion_matrix(y_test, y_pred)
+        
+        wandb.log({
+            "classification_report": class_report,
+            "confusion_matrix": conf_matrix
+        })
+        
+        results[model_name] = {
+            "clf_report": class_report,
+            "conf_matrix": conf_matrix,
+            "accuracy": accuracy,
+            "roc_auc": roc_auc,
+            "f1_score": f1_metric
+        }
+        
+        # End W&B run for this model
+        run.finish()
+
+    return results
+
+
+def json_convert(input_dict: Dict[str, Any], project: str) -> str:
+    """
+    Converts a dictionary into a JSON file and saves it to a specified directory.
+
+    Args:
+        input_dict (dict): The dictionary to be converted into a JSON file.
+        project (str): The name of the project for directory organization.
+
+    Returns:
+        str: The file path where the JSON file is saved.
+    """
+    # Ensure the folder exists
+    folder_path = f"../data/{project}/results/"
+    os.makedirs(folder_path, exist_ok=True)
+
+    file_name = f"{project}_supervized_v{random.randint(1, 100)}.json"
+    file_path = os.path.join(folder_path, file_name)
+
+    with open(file_path, 'w') as json_file:
+        json.dump(input_dict, json_file, indent=4)
+    
+    print(f"Results saved to {file_path}")
+    
+    return file_path
+
+
+def main():
+    device = "mps" if torch.backends.mps.is_available() else "cpu"
+    
+    print("Initializing project...")
+    X_train, X_test, y_train, y_test, models = initialize_project(project, version)
+    
+    print("Training classifiers...")
+    clf_train_results = training_clf(X_train, X_test, y_train, y_test, models, project, version)
+    
+    print("Saving results to JSON...")
+    json_convert(clf_train_results, project)
+    
+    print("Finished.")
+
+
+if __name__ == '__main__':
+    main()
+

data/raw/customer_profile_eda_normalization.ipynb

@@ -0,0 +1,332 @@
+{
+ "cells": [
+  {
+   "cell_type": "code",
+   "execution_count": null,
+   "metadata": {},
+   "outputs": [],
+   "source": [
+    "import pandas as pd\n"
+   ]
+  },
+  {
+   "cell_type": "code",
+   "execution_count": null,
+   "metadata": {},
+   "outputs": [],
+   "source": [
+    "from datasets import load_dataset\n",
+    "df_raw = load_dataset('Ezi/medical_and_legislators_synthetic')\n",
+    "df_pandas = df_raw['train'].to_pandas()\n",
+    "df_pandas.columns"
+   ]
+  },
+  {
+   "cell_type": "code",
+   "execution_count": null,
+   "metadata": {},
+   "outputs": [],
+   "source": [
+    "df_raw = pd.read_csv('/Users/nullzero/Documents/repos/github.com/privacy-identity/vda-simulation-medical/vda-sim-medical/data/raw/marketing_campaign.csv', sep='\\t')\n",
+    "df_raw.shape\n",
+    "df_raw[\"Response\"].unique()\n",
+    "df_raw.rename(columns={'Reponse': 'target'}, inplace=True)"
+   ]
+  },
+  {
+   "cell_type": "code",
+   "execution_count": null,
+   "metadata": {},
+   "outputs": [],
+   "source": [
+    "df = df_raw.copy()"
+   ]
+  },
+  {
+   "cell_type": "code",
+   "execution_count": null,
+   "metadata": {},
+   "outputs": [],
+   "source": [
+    "print(df.describe) "
+   ]
+  },
+  {
+   "cell_type": "code",
+   "execution_count": 122,
+   "metadata": {},
+   "outputs": [
+    {
+     "name": "stdout",
+     "output_type": "stream",
+     "text": [
+      "shape is (2240, 29) and there are ID                      0\n",
+      "Year_Birth              0\n",
+      "Education               0\n",
+      "Marital_Status          0\n",
+      "Income                 24\n",
+      "Kidhome                 0\n",
+      "Teenhome                0\n",
+      "Dt_Customer             0\n",
+      "Recency                 0\n",
+      "MntWines                0\n",
+      "MntFruits               0\n",
+      "MntMeatProducts         0\n",
+      "MntFishProducts         0\n",
+      "MntSweetProducts        0\n",
+      "MntGoldProds            0\n",
+      "NumDealsPurchases       0\n",
+      "NumWebPurchases         0\n",
+      "NumCatalogPurchases     0\n",
+      "NumStorePurchases       0\n",
+      "NumWebVisitsMonth       0\n",
+      "AcceptedCmp3            0\n",
+      "AcceptedCmp4            0\n",
+      "AcceptedCmp5            0\n",
+      "AcceptedCmp1            0\n",
+      "AcceptedCmp2            0\n",
+      "Complain                0\n",
+      "Z_CostContact           0\n",
+      "Z_Revenue               0\n",
+      "Response                0\n",
+      "dtype: int64 NAs\n",
+      "Index(['ID', 'Year_Birth', 'Education', 'Marital_Status', 'Income', 'Kidhome',\n",
+      "       'Teenhome', 'Dt_Customer', 'Recency', 'MntWines', 'MntFruits',\n",
+      "       'MntMeatProducts', 'MntFishProducts', 'MntSweetProducts',\n",
+      "       'MntGoldProds', 'NumDealsPurchases', 'NumWebPurchases',\n",
+      "       'NumCatalogPurchases', 'NumStorePurchases', 'NumWebVisitsMonth',\n",
+      "       'AcceptedCmp3', 'AcceptedCmp4', 'AcceptedCmp5', 'AcceptedCmp1',\n",
+      "       'AcceptedCmp2', 'Complain', 'Z_CostContact', 'Z_Revenue', 'Response'],\n",
+      "      dtype='object')\n"
+     ]
+    }
+   ],
+   "source": [
+    "print(f\"shape is {df.shape} and there are {df.isna().sum()} NAs\")\n",
+    "print(df.columns)\n",
+    "#columns = [\"Unnamed: 0\", \"suffix\", \"nickname, \"]"
+   ]
+  },
+  {
+   "cell_type": "code",
+   "execution_count": 124,
+   "metadata": {},
+   "outputs": [],
+   "source": [
+    "df.to_csv(\"./customer_profile_marketing.csv\")"
+   ]
+  },
+  {
+   "cell_type": "code",
+   "execution_count": null,
+   "metadata": {},
+   "outputs": [],
+   "source": [
+    "df.head(10)"
+   ]
+  },
+  {
+   "cell_type": "code",
+   "execution_count": null,
+   "metadata": {},
+   "outputs": [],
+   "source": [
+    "from sklearn.model_selection import train_test_split\n",
+    "from sklearn.cluster import KMeans, AgglomerativeClustering\n",
+    "from sklearn.feature_extraction.text import TfidfVectorizer\n",
+    "from sklearn.preprocessing import StandardScaler, OneHotEncoder\n",
+    "import seaborn as sns\n"
+   ]
+  },
+  {
+   "cell_type": "code",
+   "execution_count": null,
+   "metadata": {},
+   "outputs": [],
+   "source": [
+    "import pandas as pd\n",
+    "import numpy as np\n",
+    "import seaborn as sns\n",
+    "import matplotlib.pyplot as plt\n",
+    "from datetime import datetime\n",
+    "\n",
+    "# Drop irrelevant columns\n",
+    "df_bak = df.copy()\n",
+    "df = df_bak\n",
+    "\n",
+    "# Handle missing values\n",
+    "df['middle_name'].fillna('None', inplace=True)\n",
+    "df['suffix'].fillna('None', inplace=True)\n",
+    "\n",
+    "# Feature engineering - calculate age\n",
+    "df['birthday'] = pd.to_datetime(df['birthday'])\n",
+    "#df['age'] = (pd.Timestamp.now() - df['birthday']).astype('<m8[ns]')\n",
+    "\n",
+    "# Categorical encoding\n",
+    "df = pd.get_dummies(df, columns=['gender', 'party', 'state', 'type'])\n",
+    "\n",
+    "# Drop columns with all missing values (like washington_post_id.1)\n",
+    "df.dropna(axis=1, how='all', inplace=True)\n",
+    "\n",
+    "# Chart age before normalization\n",
+    "sns.boxplot(x=df['age'])\n",
+    "plt.title('Boxplot of Age to Detect Outliers')\n",
+    "plt.show()\n",
+    "\n",
+    "# Remove outliers for the age column (optional, if outliers are detected)\n",
+    "q_low = df['age'].quantile(0.1)\n",
+    "q_high = df['age'].quantile(0.90)\n",
+    "#df_filtered = df[(df['age'] > q_low) & (df['age'] < q_high)]\n",
+    "\n",
+    "# Visualization - Distribution of Age with limited bins\n",
+    "sns.histplot(df['age'])  # Reduced bins for better performance\n",
+    "plt.title('Distribution of Age')\n",
+    "plt.xlabel('Age')\n",
+    "plt.ylabel('Frequency')\n",
+    "plt.show()\n",
+    "\n",
+    "\n",
+    "# Normalize continuous variables\n"
+   ]
+  },
+  {
+   "cell_type": "code",
+   "execution_count": null,
+   "metadata": {},
+   "outputs": [],
+   "source": [
+    "import pandas as pd\n",
+    "import numpy as np\n",
+    "from sklearn.preprocessing import StandardScaler, LabelEncoder\n",
+    "from datetime import datetime\n",
+    "\n",
+    "# Drop irrelevant columns and duplicates\n",
+    "critical_columns = ['last_name', 'first_name', 'age', 'gender', 'state', 'district', \\\n",
+    "                    'party', 'twitter', 'facebook', 'youtube']\n",
+    "\n",
+    "# Handle missing values\n",
+    "df['middle_name'].fillna('None', inplace=True)\n",
+    "df['suffix'].fillna('None', inplace=True)\n",
+    "print(\"Suffix and middle name filled with 'None'\")\n",
+    "\n",
+    "# Feature engineering - calculate age\n",
+    "\n",
+    "# Ensure 'birthday' is in datetime format\n",
+    "df['birthday'] = pd.to_datetime(df['birthday']).dt.floor('D')\n",
+    "\n",
+    "# Calculate the current date\n",
+    "today = pd.Timestamp.now().normalize()\n",
+    "\n",
+    "# Calculate age in years\n",
+    "df['age'] = today.year - df['birthday'].dt.year\n",
+    "print(df['age'][0:3])\n",
+    "# Select only critical columns\n",
+    "df_reduced = df.copy()\n",
+    "df_reduced = df[critical_columns]\n",
+    "print(df_reduced.columns)\n"
+   ]
+  },
+  {
+   "cell_type": "code",
+   "execution_count": null,
+   "metadata": {},
+   "outputs": [],
+   "source": [
+    "\n",
+    "# Categorical encoding with LabelEncoder for binary categories\n",
+    "label_encoder = LabelEncoder()\n",
+    "df_reduced['gender'] = label_encoder.fit_transform(df_reduced['gender'])\n",
+    "\n",
+    "# Categorical encoding with OneHotEncoder for multi-class categories\n",
+    "df_reduced = pd.get_dummies(df_reduced, columns=['state', 'party', 'district'])\n",
+    "\n",
+    "# Handle online presence as binary features (e.g., presence/absence of account)\n",
+    "df_reduced['twitter'] = df_reduced['twitter'].notna().astype(int)\n",
+    "df_reduced['facebook'] = df_reduced['facebook'].notna().astype(int)\n",
+    "df_reduced['youtube'] = df_reduced['youtube'].notna().astype(int)\n",
+    "\n",
+    "# Normalize continuous variables\n",
+    "scaler = StandardScaler()\n",
+    "df_reduced[['age']] = scaler.fit_transform(df_reduced[['age']])\n",
+    "\n",
+    "# Visualization of the reduced dataframe\n",
+    "print(\"Reduced DataFrame:\")\n",
+    "print(df_reduced.head())\n",
+    "\n",
+    "# Correlation matrix to understand the relationship between features\n",
+    "import seaborn as sns\n",
+    "import matplotlib.pyplot as plt\n"
+   ]
+  },
+  {
+   "cell_type": "code",
+   "execution_count": null,
+   "metadata": {},
+   "outputs": [],
+   "source": [
+    "\n",
+    "# Missing data visualization\n",
+    "plt.figure(figsize=(12,8))\n",
+    "sns.heatmap(df.isnull(), cbar=False, cmap='viridis')\n",
+    "plt.title('Missing Data Heatmap')\n",
+    "plt.show()\n",
+    "\n",
+    "# Categorical distribution of gender\n",
+    "sns.countplot(x='gender_M', data=df)\n",
+    "plt.title('Distribution of Gender')\n",
+    "plt.xlabel('Gender (1 = Male, 0 = Female)')\n",
+    "plt.ylabel('Count')\n",
+    "plt.show()\n",
+    "\n",
+    "# Distribution by Party\n",
+    "sns.countplot(x='party', data=df)\n",
+    "plt.title('Distribution by Party')\n",
+    "plt.xlabel('Party')\n",
+    "plt.ylabel('Count')\n",
+    "plt.show()\n",
+    "\n",
+    "plt.figure(figsize=(12,8))\n",
+    "sns.heatmap(df.corr(), annot=True, fmt='.2f', cmap='coolwarm')\n",
+    "plt.title('Correlation Matrix')\n",
+    "plt.show()\n"
+   ]
+  },
+  {
+   "cell_type": "code",
+   "execution_count": null,
+   "metadata": {},
+   "outputs": [],
+   "source": [
+    "\n"
+   ]
+  },
+  {
+   "cell_type": "code",
+   "execution_count": null,
+   "metadata": {},
+   "outputs": [],
+   "source": []
+  }
+ ],
+ "metadata": {
+  "kernelspec": {
+   "display_name": "Python 3 (ipykernel)",
+   "language": "python",
+   "name": "python3"
+  },
+  "language_info": {
+   "codemirror_mode": {
+    "name": "ipython",
+    "version": 3
+   },
+   "file_extension": ".py",
+   "mimetype": "text/x-python",
+   "name": "python",
+   "nbconvert_exporter": "python",
+   "pygments_lexer": "ipython3",
+   "version": "3.11.9"
+  }
+ },
+ "nbformat": 4,
+ "nbformat_minor": 2
+}

data/raw/data_notes.md

@@ -0,0 +1,11 @@
+# Verida + PyTorch Differential Privacy Experiment
+
+# Dataset: 
+https://huggingface.co/datasets/Ezi/medical_and_legislators_synthetic
+
+# Objective:
+
+#
+
+# Notes:
+1. Appended with synthetically generated VDA DiD

data/raw/hf/.gitattributes

@@ -0,0 +1,58 @@
+*.7z filter=lfs diff=lfs merge=lfs -text
+*.arrow filter=lfs diff=lfs merge=lfs -text
+*.bin filter=lfs diff=lfs merge=lfs -text
+*.bz2 filter=lfs diff=lfs merge=lfs -text
+*.ckpt filter=lfs diff=lfs merge=lfs -text
+*.ftz filter=lfs diff=lfs merge=lfs -text
+*.gz filter=lfs diff=lfs merge=lfs -text
+*.h5 filter=lfs diff=lfs merge=lfs -text
+*.joblib filter=lfs diff=lfs merge=lfs -text
+*.lfs.* filter=lfs diff=lfs merge=lfs -text
+*.lz4 filter=lfs diff=lfs merge=lfs -text
+*.mlmodel filter=lfs diff=lfs merge=lfs -text
+*.model filter=lfs diff=lfs merge=lfs -text
+*.msgpack filter=lfs diff=lfs merge=lfs -text
+*.npy filter=lfs diff=lfs merge=lfs -text
+*.npz filter=lfs diff=lfs merge=lfs -text
+*.onnx filter=lfs diff=lfs merge=lfs -text
+*.ot filter=lfs diff=lfs merge=lfs -text
+*.parquet filter=lfs diff=lfs merge=lfs -text
+*.pb filter=lfs diff=lfs merge=lfs -text
+*.pickle filter=lfs diff=lfs merge=lfs -text
+*.pkl filter=lfs diff=lfs merge=lfs -text
+*.pt filter=lfs diff=lfs merge=lfs -text
+*.pth filter=lfs diff=lfs merge=lfs -text
+*.rar filter=lfs diff=lfs merge=lfs -text
+*.safetensors filter=lfs diff=lfs merge=lfs -text
+saved_model/**/* filter=lfs diff=lfs merge=lfs -text
+*.tar.* filter=lfs diff=lfs merge=lfs -text
+*.tar filter=lfs diff=lfs merge=lfs -text
+*.tflite filter=lfs diff=lfs merge=lfs -text
+*.tgz filter=lfs diff=lfs merge=lfs -text
+*.wasm filter=lfs diff=lfs merge=lfs -text
+*.xz filter=lfs diff=lfs merge=lfs -text
+*.zip filter=lfs diff=lfs merge=lfs -text
+*.zst filter=lfs diff=lfs merge=lfs -text
+*tfevents* filter=lfs diff=lfs merge=lfs -text
+# Audio files - uncompressed
+*.pcm filter=lfs diff=lfs merge=lfs -text
+*.sam filter=lfs diff=lfs merge=lfs -text
+*.raw filter=lfs diff=lfs merge=lfs -text
+# Audio files - compressed
+*.aac filter=lfs diff=lfs merge=lfs -text
+*.flac filter=lfs diff=lfs merge=lfs -text
+*.mp3 filter=lfs diff=lfs merge=lfs -text
+*.ogg filter=lfs diff=lfs merge=lfs -text
+*.wav filter=lfs diff=lfs merge=lfs -text
+# Image files - uncompressed
+*.bmp filter=lfs diff=lfs merge=lfs -text
+*.gif filter=lfs diff=lfs merge=lfs -text
+*.png filter=lfs diff=lfs merge=lfs -text
+*.tiff filter=lfs diff=lfs merge=lfs -text
+# Image files - compressed
+*.jpg filter=lfs diff=lfs merge=lfs -text
+*.jpeg filter=lfs diff=lfs merge=lfs -text
+*.webp filter=lfs diff=lfs merge=lfs -text
+# Video files - compressed
+*.mp4 filter=lfs diff=lfs merge=lfs -text
+*.webm filter=lfs diff=lfs merge=lfs -text

data/raw/hf/README.md

@@ -0,0 +1,55 @@
+---
+annotations_creators:
+- machine-generated
+language:
+- en
+language_creators:
+- machine-generated
+license:
+- mit
+multilinguality:
+- monolingual
+pretty_name: Medical data with DiD's randomly generated
+size_categories:
+- n<1K
+source_datasets:
+- original
+tags: []
+task_categories:
+- tabular-classification
+- text-classification
+task_ids:
+- tabular-multi-class-classification
+- multi-class-classification
+---
+
+# Dataset Card for [Dataset Name]
+
+## Table of Contents
+- [Table of Contents](#table-of-contents)
+- [Dataset Description](#dataset-description)
+  - [Dataset Summary](#dataset-summary)
+  - [Supported Tasks and Leaderboards](#supported-tasks-and-leaderboards)
+  - [Languages](#languages)
+- [Dataset Structure](#dataset-structure)
+  - [Data Instances](#data-instances)
+  - [Data Fields](#data-fields)
+  - [Data Splits](#data-splits)
+- [Dataset Creation](#dataset-creation)
+  - [Curation Rationale](#curation-rationale)
+  - [Source Data](#source-data)
+  - [Annotations](#annotations)
+  - [Personal and Sensitive Information](#personal-and-sensitive-information)
+- [Considerations for Using the Data](#considerations-for-using-the-data)
+  - [Social Impact of Dataset](#social-impact-of-dataset)
+  - [Discussion of Biases](#discussion-of-biases)
+  - [Other Known Limitations](#other-known-limitations)
+- [Additional Information](#additional-information)
+  - [Dataset Curators](#dataset-curators)
+  - [Licensing Information](#licensing-information)
+  - [Citation Information](#citation-information)
+  - [Contributions](#contributions)
+
+## Dataset Description
+Medical data with DiD's for differential privacy.
+

data/raw/medical_data_raw.ipynb

@@ -0,0 +1,207 @@
+{
+ "cells": [
+  {
+   "cell_type": "markdown",
+   "metadata": {},
+   "source": [
+    "# PyTorch Differential Privacy Experiment"
+   ]
+  },
+  {
+   "cell_type": "code",
+   "execution_count": 1,
+   "metadata": {},
+   "outputs": [
+    {
+     "name": "stdout",
+     "output_type": "stream",
+     "text": [
+      "Collecting torch\n",
+      "  Downloading torch-2.4.0-cp311-none-macosx_11_0_arm64.whl.metadata (26 kB)\n",
+      "Collecting torchvision\n",
+      "  Downloading torchvision-0.19.0-cp311-cp311-macosx_11_0_arm64.whl.metadata (6.0 kB)\n",
+      "Collecting opacus\n",
+      "  Downloading opacus-1.5.2-py3-none-any.whl.metadata (7.9 kB)\n",
+      "Requirement already satisfied: numpy in /Users/nullzero/Library/Caches/pypoetry/virtualenvs/verida-differential-privacy-OB45ac0m-py3.11/lib/python3.11/site-packages (2.1.0)\n",
+      "Requirement already satisfied: pandas in /Users/nullzero/Library/Caches/pypoetry/virtualenvs/verida-differential-privacy-OB45ac0m-py3.11/lib/python3.11/site-packages (2.2.2)\n",
+      "Requirement already satisfied: filelock in /Users/nullzero/Library/Caches/pypoetry/virtualenvs/verida-differential-privacy-OB45ac0m-py3.11/lib/python3.11/site-packages (from torch) (3.15.4)\n",
+      "Requirement already satisfied: typing-extensions>=4.8.0 in /Users/nullzero/Library/Caches/pypoetry/virtualenvs/verida-differential-privacy-OB45ac0m-py3.11/lib/python3.11/site-packages (from torch) (4.12.2)\n",
+      "Collecting sympy (from torch)\n",
+      "  Downloading sympy-1.13.2-py3-none-any.whl.metadata (12 kB)\n",
+      "Collecting networkx (from torch)\n",
+      "  Using cached networkx-3.3-py3-none-any.whl.metadata (5.1 kB)\n",
+      "Requirement already satisfied: jinja2 in /Users/nullzero/Library/Caches/pypoetry/virtualenvs/verida-differential-privacy-OB45ac0m-py3.11/lib/python3.11/site-packages (from torch) (3.1.4)\n",
+      "Requirement already satisfied: fsspec in /Users/nullzero/Library/Caches/pypoetry/virtualenvs/verida-differential-privacy-OB45ac0m-py3.11/lib/python3.11/site-packages (from torch) (2024.6.1)\n",
+      "Collecting pillow!=8.3.*,>=5.3.0 (from torchvision)\n",
+      "  Downloading pillow-10.4.0-cp311-cp311-macosx_11_0_arm64.whl.metadata (9.2 kB)\n",
+      "Collecting numpy\n",
+      "  Using cached numpy-1.26.4-cp311-cp311-macosx_11_0_arm64.whl.metadata (114 kB)\n",
+      "Collecting scipy>=1.2 (from opacus)\n",
+      "  Downloading scipy-1.14.1-cp311-cp311-macosx_14_0_arm64.whl.metadata (60 kB)\n",
+      "Collecting opt-einsum>=3.3.0 (from opacus)\n",
+      "  Downloading opt_einsum-3.3.0-py3-none-any.whl.metadata (6.5 kB)\n",
+      "Requirement already satisfied: python-dateutil>=2.8.2 in /Users/nullzero/Library/Caches/pypoetry/virtualenvs/verida-differential-privacy-OB45ac0m-py3.11/lib/python3.11/site-packages (from pandas) (2.9.0.post0)\n",
+      "Requirement already satisfied: pytz>=2020.1 in /Users/nullzero/Library/Caches/pypoetry/virtualenvs/verida-differential-privacy-OB45ac0m-py3.11/lib/python3.11/site-packages (from pandas) (2024.1)\n",
+      "Requirement already satisfied: tzdata>=2022.7 in /Users/nullzero/Library/Caches/pypoetry/virtualenvs/verida-differential-privacy-OB45ac0m-py3.11/lib/python3.11/site-packages (from pandas) (2024.1)\n",
+      "Requirement already satisfied: six>=1.5 in /Users/nullzero/Library/Caches/pypoetry/virtualenvs/verida-differential-privacy-OB45ac0m-py3.11/lib/python3.11/site-packages (from python-dateutil>=2.8.2->pandas) (1.16.0)\n",
+      "Requirement already satisfied: MarkupSafe>=2.0 in /Users/nullzero/Library/Caches/pypoetry/virtualenvs/verida-differential-privacy-OB45ac0m-py3.11/lib/python3.11/site-packages (from jinja2->torch) (2.1.5)\n",
+      "Collecting mpmath<1.4,>=1.1.0 (from sympy->torch)\n",
+      "  Using cached mpmath-1.3.0-py3-none-any.whl.metadata (8.6 kB)\n",
+      "Downloading torch-2.4.0-cp311-none-macosx_11_0_arm64.whl (62.1 MB)\n",
+      "\u001b[2K   \u001b[38;2;114;156;31m━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━\u001b[0m \u001b[32m62.1/62.1 MB\u001b[0m \u001b[31m440.8 kB/s\u001b[0m eta \u001b[36m0:00:00\u001b[0mm eta \u001b[36m0:00:01\u001b[0m[36m0:00:04\u001b[0m\n",
+      "\u001b[?25hDownloading torchvision-0.19.0-cp311-cp311-macosx_11_0_arm64.whl (1.7 MB)\n",
+      "\u001b[2K   \u001b[38;2;114;156;31m━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━\u001b[0m \u001b[32m1.7/1.7 MB\u001b[0m \u001b[31m560.7 kB/s\u001b[0m eta \u001b[36m0:00:00\u001b[0m1m557.0 kB/s\u001b[0m eta \u001b[36m0:00:01\u001b[0m\n",
+      "\u001b[?25hDownloading opacus-1.5.2-py3-none-any.whl (239 kB)\n",
+      "Using cached numpy-1.26.4-cp311-cp311-macosx_11_0_arm64.whl (14.0 MB)\n",
+      "Downloading opt_einsum-3.3.0-py3-none-any.whl (65 kB)\n",
+      "Downloading pillow-10.4.0-cp311-cp311-macosx_11_0_arm64.whl (3.4 MB)\n",
+      "\u001b[2K   \u001b[38;2;114;156;31m━━━━━━━━━━━━━━━━━━━━━━━━��━━━━━━━━━━━━━━━\u001b[0m \u001b[32m3.4/3.4 MB\u001b[0m \u001b[31m457.2 kB/s\u001b[0m eta \u001b[36m0:00:00\u001b[0m kB/s\u001b[0m eta \u001b[36m0:00:01\u001b[0m:02\u001b[0m\n",
+      "\u001b[?25hDownloading scipy-1.14.1-cp311-cp311-macosx_14_0_arm64.whl (23.1 MB)\n",
+      "\u001b[2K   \u001b[38;2;114;156;31m━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━\u001b[0m \u001b[32m23.1/23.1 MB\u001b[0m \u001b[31m442.0 kB/s\u001b[0m eta \u001b[36m0:00:00\u001b[0mm eta \u001b[36m0:00:01\u001b[0m[36m0:00:02\u001b[0m\n",
+      "\u001b[?25hUsing cached networkx-3.3-py3-none-any.whl (1.7 MB)\n",
+      "Downloading sympy-1.13.2-py3-none-any.whl (6.2 MB)\n",
+      "\u001b[2K   \u001b[38;2;114;156;31m━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━\u001b[0m \u001b[32m6.2/6.2 MB\u001b[0m \u001b[31m462.5 kB/s\u001b[0m eta \u001b[36m0:00:00\u001b[0m1m487.3 kB/s\u001b[0m eta \u001b[36m0:00:01\u001b[0m\n",
+      "\u001b[?25hUsing cached mpmath-1.3.0-py3-none-any.whl (536 kB)\n",
+      "Installing collected packages: mpmath, sympy, pillow, numpy, networkx, torch, scipy, opt-einsum, torchvision, opacus\n",
+      "  Attempting uninstall: numpy\n",
+      "    Found existing installation: numpy 2.1.0\n",
+      "    Uninstalling numpy-2.1.0:\n",
+      "      Successfully uninstalled numpy-2.1.0\n",
+      "Successfully installed mpmath-1.3.0 networkx-3.3 numpy-1.26.4 opacus-1.5.2 opt-einsum-3.3.0 pillow-10.4.0 scipy-1.14.1 sympy-1.13.2 torch-2.4.0 torchvision-0.19.0\n",
+      "Note: you may need to restart the kernel to use updated packages.\n",
+      "Requirement already satisfied: wandb in /Users/nullzero/Library/Caches/pypoetry/virtualenvs/verida-differential-privacy-OB45ac0m-py3.11/lib/python3.11/site-packages (0.17.7)\n",
+      "Requirement already satisfied: datasets in /Users/nullzero/Library/Caches/pypoetry/virtualenvs/verida-differential-privacy-OB45ac0m-py3.11/lib/python3.11/site-packages (2.21.0)\n",
+      "Requirement already satisfied: tqdm in /Users/nullzero/Library/Caches/pypoetry/virtualenvs/verida-differential-privacy-OB45ac0m-py3.11/lib/python3.11/site-packages (4.66.5)\n",
+      "Requirement already satisfied: click!=8.0.0,>=7.1 in /Users/nullzero/Library/Caches/pypoetry/virtualenvs/verida-differential-privacy-OB45ac0m-py3.11/lib/python3.11/site-packages (from wandb) (8.1.7)\n",
+      "Requirement already satisfied: docker-pycreds>=0.4.0 in /Users/nullzero/Library/Caches/pypoetry/virtualenvs/verida-differential-privacy-OB45ac0m-py3.11/lib/python3.11/site-packages (from wandb) (0.4.0)\n",
+      "Requirement already satisfied: gitpython!=3.1.29,>=1.0.0 in /Users/nullzero/Library/Caches/pypoetry/virtualenvs/verida-differential-privacy-OB45ac0m-py3.11/lib/python3.11/site-packages (from wandb) (3.1.43)\n",
+      "Requirement already satisfied: platformdirs in /Users/nullzero/Library/Caches/pypoetry/virtualenvs/verida-differential-privacy-OB45ac0m-py3.11/lib/python3.11/site-packages (from wandb) (4.2.2)\n",
+      "Requirement already satisfied: protobuf!=4.21.0,<6,>=3.19.0 in /Users/nullzero/Library/Caches/pypoetry/virtualenvs/verida-differential-privacy-OB45ac0m-py3.11/lib/python3.11/site-packages (from wandb) (5.27.3)\n",
+      "Requirement already satisfied: psutil>=5.0.0 in /Users/nullzero/Library/Caches/pypoetry/virtualenvs/verida-differential-privacy-OB45ac0m-py3.11/lib/python3.11/site-packages (from wandb) (6.0.0)\n",
+      "Requirement already satisfied: pyyaml in /Users/nullzero/Library/Caches/pypoetry/virtualenvs/verida-differential-privacy-OB45ac0m-py3.11/lib/python3.11/site-packages (from wandb) (6.0.2)\n",
+      "Requirement already satisfied: requests<3,>=2.0.0 in /Users/nullzero/Library/Caches/pypoetry/virtualenvs/verida-differential-privacy-OB45ac0m-py3.11/lib/python3.11/site-packages (from wandb) (2.32.3)\n",
+      "Requirement already satisfied: sentry-sdk>=1.0.0 in /Users/nullzero/Library/Caches/pypoetry/virtualenvs/verida-differential-privacy-OB45ac0m-py3.11/lib/python3.11/site-packages (from wandb) (2.13.0)\n",
+      "Requirement already satisfied: setproctitle in /Users/nullzero/Library/Caches/pypoetry/virtualenvs/verida-differential-privacy-OB45ac0m-py3.11/lib/python3.11/site-packages (from wandb) (1.3.3)\n",
+      "Requirement already satisfied: setuptools in /Users/nullzero/Library/Caches/pypoetry/virtualenvs/verida-differential-privacy-OB45ac0m-py3.11/lib/python3.11/site-packages (from wandb) (73.0.1)\n",
+      "Requirement already satisfied: filelock in /Users/nullzero/Library/Caches/pypoetry/virtualenvs/verida-differential-privacy-OB45ac0m-py3.11/lib/python3.11/site-packages (from datasets) (3.15.4)\n",
+      "Requirement already satisfied: numpy>=1.17 in /Users/nullzero/Library/Caches/pypoetry/virtualenvs/verida-differential-privacy-OB45ac0m-py3.11/lib/python3.11/site-packages (from datasets) (1.26.4)\n",
+      "Requirement already satisfied: pyarrow>=15.0.0 in /Users/nullzero/Library/Caches/pypoetry/virtualenvs/verida-differential-privacy-OB45ac0m-py3.11/lib/python3.11/site-packages (from datasets) (17.0.0)\n",
+      "Requirement already satisfied: dill<0.3.9,>=0.3.0 in /Users/nullzero/Library/Caches/pypoetry/virtualenvs/verida-differential-privacy-OB45ac0m-py3.11/lib/python3.11/site-packages (from datasets) (0.3.8)\n",
+      "Requirement already satisfied: pandas in /Users/nullzero/Library/Caches/pypoetry/virtualenvs/verida-differential-privacy-OB45ac0m-py3.11/lib/python3.11/site-packages (from datasets) (2.2.2)\n",
+      "Requirement already satisfied: xxhash in /Users/nullzero/Library/Caches/pypoetry/virtualenvs/verida-differential-privacy-OB45ac0m-py3.11/lib/python3.11/site-packages (from datasets) (3.5.0)\n",
+      "Requirement already satisfied: multiprocess in /Users/nullzero/Library/Caches/pypoetry/virtualenvs/verida-differential-privacy-OB45ac0m-py3.11/lib/python3.11/site-packages (from datasets) (0.70.16)\n",
+      "Requirement already satisfied: fsspec<=2024.6.1,>=2023.1.0 in /Users/nullzero/Library/Caches/pypoetry/virtualenvs/verida-differential-privacy-OB45ac0m-py3.11/lib/python3.11/site-packages (from fsspec[http]<=2024.6.1,>=2023.1.0->datasets) (2024.6.1)\n",
+      "Requirement already satisfied: aiohttp in /Users/nullzero/Library/Caches/pypoetry/virtualenvs/verida-differential-privacy-OB45ac0m-py3.11/lib/python3.11/site-packages (from datasets) (3.10.5)\n",
+      "Requirement already satisfied: huggingface-hub>=0.21.2 in /Users/nullzero/Library/Caches/pypoetry/virtualenvs/verida-differential-privacy-OB45ac0m-py3.11/lib/python3.11/site-packages (from datasets) (0.24.6)\n",
+      "Requirement already satisfied: packaging in /Users/nullzero/Library/Caches/pypoetry/virtualenvs/verida-differential-privacy-OB45ac0m-py3.11/lib/python3.11/site-packages (from datasets) (24.1)\n",
+      "Requirement already satisfied: six>=1.4.0 in /Users/nullzero/Library/Caches/pypoetry/virtualenvs/verida-differential-privacy-OB45ac0m-py3.11/lib/python3.11/site-packages (from docker-pycreds>=0.4.0->wandb) (1.16.0)\n",
+      "Requirement already satisfied: aiohappyeyeballs>=2.3.0 in /Users/nullzero/Library/Caches/pypoetry/virtualenvs/verida-differential-privacy-OB45ac0m-py3.11/lib/python3.11/site-packages (from aiohttp->datasets) (2.4.0)\n",
+      "Requirement already satisfied: aiosignal>=1.1.2 in /Users/nullzero/Library/Caches/pypoetry/virtualenvs/verida-differential-privacy-OB45ac0m-py3.11/lib/python3.11/site-packages (from aiohttp->datasets) (1.3.1)\n",
+      "Requirement already satisfied: attrs>=17.3.0 in /Users/nullzero/Library/Caches/pypoetry/virtualenvs/verida-differential-privacy-OB45ac0m-py3.11/lib/python3.11/site-packages (from aiohttp->datasets) (24.2.0)\n",
+      "Requirement already satisfied: frozenlist>=1.1.1 in /Users/nullzero/Library/Caches/pypoetry/virtualenvs/verida-differential-privacy-OB45ac0m-py3.11/lib/python3.11/site-packages (from aiohttp->datasets) (1.4.1)\n",
+      "Requirement already satisfied: multidict<7.0,>=4.5 in /Users/nullzero/Library/Caches/pypoetry/virtualenvs/verida-differential-privacy-OB45ac0m-py3.11/lib/python3.11/site-packages (from aiohttp->datasets) (6.0.5)\n",
+      "Requirement already satisfied: yarl<2.0,>=1.0 in /Users/nullzero/Library/Caches/pypoetry/virtualenvs/verida-differential-privacy-OB45ac0m-py3.11/lib/python3.11/site-packages (from aiohttp->datasets) (1.9.4)\n",
+      "Requirement already satisfied: gitdb<5,>=4.0.1 in /Users/nullzero/Library/Caches/pypoetry/virtualenvs/verida-differential-privacy-OB45ac0m-py3.11/lib/python3.11/site-packages (from gitpython!=3.1.29,>=1.0.0->wandb) (4.0.11)\n",
+      "Requirement already satisfied: typing-extensions>=3.7.4.3 in /Users/nullzero/Library/Caches/pypoetry/virtualenvs/verida-differential-privacy-OB45ac0m-py3.11/lib/python3.11/site-packages (from huggingface-hub>=0.21.2->datasets) (4.12.2)\n",
+      "Requirement already satisfied: charset-normalizer<4,>=2 in /Users/nullzero/Library/Caches/pypoetry/virtualenvs/verida-differential-privacy-OB45ac0m-py3.11/lib/python3.11/site-packages (from requests<3,>=2.0.0->wandb) (3.3.2)\n",
+      "Requirement already satisfied: idna<4,>=2.5 in /Users/nullzero/Library/Caches/pypoetry/virtualenvs/verida-differential-privacy-OB45ac0m-py3.11/lib/python3.11/site-packages (from requests<3,>=2.0.0->wandb) (3.7)\n",
+      "Requirement already satisfied: urllib3<3,>=1.21.1 in /Users/nullzero/Library/Caches/pypoetry/virtualenvs/verida-differential-privacy-OB45ac0m-py3.11/lib/python3.11/site-packages (from requests<3,>=2.0.0->wandb) (2.2.2)\n",
+      "Requirement already satisfied: certifi>=2017.4.17 in /Users/nullzero/Library/Caches/pypoetry/virtualenvs/verida-differential-privacy-OB45ac0m-py3.11/lib/python3.11/site-packages (from requests<3,>=2.0.0->wandb) (2024.7.4)\n",
+      "Requirement already satisfied: python-dateutil>=2.8.2 in /Users/nullzero/Library/Caches/pypoetry/virtualenvs/verida-differential-privacy-OB45ac0m-py3.11/lib/python3.11/site-packages (from pandas->datasets) (2.9.0.post0)\n",
+      "Requirement already satisfied: pytz>=2020.1 in /Users/nullzero/Library/Caches/pypoetry/virtualenvs/verida-differential-privacy-OB45ac0m-py3.11/lib/python3.11/site-packages (from pandas->datasets) (2024.1)\n",
+      "Requirement already satisfied: tzdata>=2022.7 in /Users/nullzero/Library/Caches/pypoetry/virtualenvs/verida-differential-privacy-OB45ac0m-py3.11/lib/python3.11/site-packages (from pandas->datasets) (2024.1)\n",
+      "Requirement already satisfied: smmap<6,>=3.0.1 in /Users/nullzero/Library/Caches/pypoetry/virtualenvs/verida-differential-privacy-OB45ac0m-py3.11/lib/python3.11/site-packages (from gitdb<5,>=4.0.1->gitpython!=3.1.29,>=1.0.0->wandb) (5.0.1)\n",
+      "Note: you may need to restart the kernel to use updated packages.\n"
+     ]
+    }
+   ],
+   "source": [
+    "%pip install -qqq torch torchvision opacus numpy pandas\n",
+    "%pip install -qqq wandb datasets tqdm\n"
+   ]
+  },
+  {
+   "cell_type": "code",
+   "execution_count": 2,
+   "metadata": {},
+   "outputs": [],
+   "source": [
+    "import os\n",
+    "import torch \n",
+    "from dotenv import load_dotenv\n",
+    "import wandb \n",
+    "import logging\n",
+    "import shutil\n",
+    "import sys\n",
+    "from datetime import datetime, timedelta\n",
+    "\n",
+    "import argparse\n",
+    "from collections import Counter\n",
+    "from pathlib import Path\n",
+    "from statistics import mean\n",
+    "\n",
+    "import torch\n",
+    "import torch.nn as nn\n",
+    "from opacus import PrivacyEngine\n",
+    "from opacus.layers import DPGRU, DPLSTM, DPRNN\n",
+    "from torch.nn.utils.rnn import pad_sequence\n",
+    "from torch.utils.data import DataLoader, Dataset\n",
+    "from tqdm import tqdm, tqdm_notebook"
+   ]
+  },
+  {
+   "cell_type": "code",
+   "execution_count": 3,
+   "metadata": {},
+   "outputs": [],
+   "source": [
+    "device = \"mps\" if torch.backends.mps.is_available() else \"cpu\"\n",
+    "if os.path.exists('.env'):\n",
+    "    load_dotenv('.env')\n"
+   ]
+  },
+  {
+   "cell_type": "code",
+   "execution_count": null,
+   "metadata": {},
+   "outputs": [],
+   "source": [
+    "logging.basicConfig(\n",
+    "    format=\"%(asctime)s:%(levelname)s:%(message)s\",\n",
+    "    datefmt=\"%m/%d/%Y %H:%M:%S\",\n",
+    "    stream=sys.stdout,\n",
+    ")\n",
+    "logger = logging.getLogger(\"ddp\")\n",
+    "logger.setLevel(level=logging.INFO)\n"
+   ]
+  },
+  {
+   "cell_type": "code",
+   "execution_count": null,
+   "metadata": {},
+   "outputs": [],
+   "source": [
+    "wandb.login(key=os.getenv('WANDB_API_KEY'))\n",
+    "wandb.init(project=\"verida-pii\", name=\"verida_data_raw\")"
+   ]
+  },
+  {
+   "cell_type": "code",
+   "execution_count": null,
+   "metadata": {},
+   "outputs": [],
+   "source": [
+    "data_name = 'Ezi/medical_and_legislators_synthetic'"
+   ]
+  }
+ ],
+ "metadata": {
+  "kernelspec": {
+   "display_name": "Python 3 (ipykernel)",
+   "language": "python",
+   "name": "python3"
+  }
+ },
+ "nbformat": 4,
+ "nbformat_minor": 2
+}

data/raw/raw_data.ipynb

@@ -0,0 +1,662 @@
+{
+ "cells": [
+  {
+   "cell_type": "markdown",
+   "metadata": {},
+   "source": [
+    "# Data Preparation  "
+   ]
+  },
+  {
+   "cell_type": "code",
+   "execution_count": 44,
+   "metadata": {},
+   "outputs": [],
+   "source": [
+    "!poetry add -qqq python-dotenv datasets wandb didkit\n"
+   ]
+  },
+  {
+   "cell_type": "code",
+   "execution_count": null,
+   "metadata": {},
+   "outputs": [],
+   "source": []
+  },
+  {
+   "cell_type": "code",
+   "execution_count": 34,
+   "metadata": {},
+   "outputs": [],
+   "source": [
+    "import os\n",
+    "from dotenv import load_dotenv, find_dotenv\n",
+    "if os.path.exists('../env'):\n",
+    "    load_dotenv(find_dotenv())\n",
+    "import wandb"
+   ]
+  },
+  {
+   "cell_type": "code",
+   "execution_count": 36,
+   "metadata": {},
+   "outputs": [
+    {
+     "name": "stderr",
+     "output_type": "stream",
+     "text": [
+      "\u001b[34m\u001b[1mwandb\u001b[0m: \u001b[33mWARNING\u001b[0m Calling wandb.login() after wandb.init() has no effect.\n"
+     ]
+    },
+    {
+     "data": {
+      "text/html": [
+       "Finishing last run (ID:pnvhnkh8) before initializing another..."
+      ],
+      "text/plain": [
+       "<IPython.core.display.HTML object>"
+      ]
+     },
+     "metadata": {},
+     "output_type": "display_data"
+    },
+    {
+     "data": {
+      "application/vnd.jupyter.widget-view+json": {
+       "model_id": "19afd929911c42f6ab4e9604948ade90",
+       "version_major": 2,
+       "version_minor": 0
+      },
+      "text/plain": [
+       "VBox(children=(Label(value='0.001 MB of 0.001 MB uploaded\\r'), FloatProgress(value=1.0, max=1.0)))"
+      ]
+     },
+     "metadata": {},
+     "output_type": "display_data"
+    },
+    {
+     "data": {
+      "text/html": [
+       " View run <strong style=\"color:#cdcd00\">verida_data_raw</strong> at: <a href='https://wandb.ai/orion-agents/verida-pii/runs/pnvhnkh8' target=\"_blank\">https://wandb.ai/orion-agents/verida-pii/runs/pnvhnkh8</a><br/> View project at: <a href='https://wandb.ai/orion-agents/verida-pii' target=\"_blank\">https://wandb.ai/orion-agents/verida-pii</a><br/>Synced 5 W&B file(s), 0 media file(s), 0 artifact file(s) and 0 other file(s)"
+      ],
+      "text/plain": [
+       "<IPython.core.display.HTML object>"
+      ]
+     },
+     "metadata": {},
+     "output_type": "display_data"
+    },
+    {
+     "data": {
+      "text/html": [
+       "Find logs at: <code>./wandb/run-20240825_035519-pnvhnkh8/logs</code>"
+      ],
+      "text/plain": [
+       "<IPython.core.display.HTML object>"
+      ]
+     },
+     "metadata": {},
+     "output_type": "display_data"
+    },
+    {
+     "data": {
+      "text/html": [
+       "The new W&B backend becomes opt-out in version 0.18.0; try it out with `wandb.require(\"core\")`! See https://wandb.me/wandb-core for more information."
+      ],
+      "text/plain": [
+       "<IPython.core.display.HTML object>"
+      ]
+     },
+     "metadata": {},
+     "output_type": "display_data"
+    },
+    {
+     "data": {
+      "text/html": [
+       "Successfully finished last run (ID:pnvhnkh8). Initializing new run:<br/>"
+      ],
+      "text/plain": [
+       "<IPython.core.display.HTML object>"
+      ]
+     },
+     "metadata": {},
+     "output_type": "display_data"
+    },
+    {
+     "data": {
+      "application/vnd.jupyter.widget-view+json": {
+       "model_id": "76495a6b8c1843b09162b9ec31d99dfe",
+       "version_major": 2,
+       "version_minor": 0
+      },
+      "text/plain": [
+       "VBox(children=(Label(value='Waiting for wandb.init()...\\r'), FloatProgress(value=0.01112003148947325, max=1.0)…"
+      ]
+     },
+     "metadata": {},
+     "output_type": "display_data"
+    },
+    {
+     "data": {
+      "text/html": [
+       "Tracking run with wandb version 0.17.7"
+      ],
+      "text/plain": [
+       "<IPython.core.display.HTML object>"
+      ]
+     },
+     "metadata": {},
+     "output_type": "display_data"
+    },
+    {
+     "data": {
+      "text/html": [
+       "Run data is saved locally in <code>/Users/nullzero/Documents/repos/github.com/privacy-identity/vda-simulation-medical/vda-sim-medical/wandb/run-20240825_035604-69f6mbdr</code>"
+      ],
+      "text/plain": [
+       "<IPython.core.display.HTML object>"
+      ]
+     },
+     "metadata": {},
+     "output_type": "display_data"
+    },
+    {
+     "data": {
+      "text/html": [
+       "Syncing run <strong><a href='https://wandb.ai/orion-agents/verida-pii/runs/69f6mbdr' target=\"_blank\">verida_data_raw</a></strong> to <a href='https://wandb.ai/orion-agents/verida-pii' target=\"_blank\">Weights & Biases</a> (<a href='https://wandb.me/run' target=\"_blank\">docs</a>)<br/>"
+      ],
+      "text/plain": [
+       "<IPython.core.display.HTML object>"
+      ]
+     },
+     "metadata": {},
+     "output_type": "display_data"
+    },
+    {
+     "data": {
+      "text/html": [
+       " View project at <a href='https://wandb.ai/orion-agents/verida-pii' target=\"_blank\">https://wandb.ai/orion-agents/verida-pii</a>"
+      ],
+      "text/plain": [
+       "<IPython.core.display.HTML object>"
+      ]
+     },
+     "metadata": {},
+     "output_type": "display_data"
+    },
+    {
+     "data": {
+      "text/html": [
+       " View run at <a href='https://wandb.ai/orion-agents/verida-pii/runs/69f6mbdr' target=\"_blank\">https://wandb.ai/orion-agents/verida-pii/runs/69f6mbdr</a>"
+      ],
+      "text/plain": [
+       "<IPython.core.display.HTML object>"
+      ]
+     },
+     "metadata": {},
+     "output_type": "display_data"
+    },
+    {
+     "name": "stderr",
+     "output_type": "stream",
+     "text": [
+      "\u001b[34m\u001b[1mwandb\u001b[0m: Network error resolved after 1:13:41.968146, resuming normal operation.\n",
+      "\u001b[34m\u001b[1mwandb\u001b[0m: Network error resolved after 0:42:49.841123, resuming normal operation.\n",
+      "\u001b[34m\u001b[1mwandb\u001b[0m: Network error resolved after 0:18:54.049113, resuming normal operation.\n"
+     ]
+    }
+   ],
+   "source": [
+    "wandb.login(key=os.getenv('WANDB_API_KEY'))\n",
+    "run = wandb.init(project=\"verida-pii\", name=\"verida_data_raw\")"
+   ]
+  },
+  {
+   "cell_type": "code",
+   "execution_count": 62,
+   "metadata": {},
+   "outputs": [
+    {
+     "name": "stdout",
+     "output_type": "stream",
+     "text": [
+      "539\n"
+     ]
+    }
+   ],
+   "source": [
+    "from datasets import load_dataset\n",
+    "import pandas as pd\n",
+    "data_name=\"Ezi/medical_and_legislators_synthetic\"\n",
+    "data = load_dataset(path=data_name, split='train')\n",
+    "data_df = data.to_pandas()\n",
+    "data_df.head()\n",
+    "print(len(data_df))\n",
+    "\n"
+   ]
+  },
+  {
+   "cell_type": "code",
+   "execution_count": 30,
+   "metadata": {},
+   "outputs": [
+    {
+     "data": {
+      "text/plain": [
+       "'mps'"
+      ]
+     },
+     "execution_count": 30,
+     "metadata": {},
+     "output_type": "execute_result"
+    }
+   ],
+   "source": [
+    "device = \"mps\" if torch.backends.mps.is_available() else \"cpu\"\n",
+    "device"
+   ]
+  },
+  {
+   "cell_type": "code",
+   "execution_count": 63,
+   "metadata": {},
+   "outputs": [],
+   "source": [
+    "# DiD Generator\n",
+    "import didkit\n",
+    "\n",
+    "def generate_did():\n",
+    "    key = didkit.generate_ed25519_key()\n",
+    "    did = didkit.key_to_did(\"key\", key)\n",
+    "    return did, key"
+   ]
+  },
+  {
+   "cell_type": "code",
+   "execution_count": 64,
+   "metadata": {},
+   "outputs": [],
+   "source": [
+    "from tqdm import tqdm, tqdm_notebook, tqdm_pandas\n",
+    "import pandas as pd"
+   ]
+  },
+  {
+   "cell_type": "code",
+   "execution_count": 70,
+   "metadata": {},
+   "outputs": [
+    {
+     "data": {
+      "text/plain": [
+       "Index(['last_name', 'first_name', 'middle_name', 'suffix', 'nickname',\n",
+       "       'full_name', 'birthday', 'gender', 'type', 'state', 'district',\n",
+       "       'senate_class', 'party', 'url', 'address', 'phone', 'contact_form',\n",
+       "       'rss_url', 'twitter', 'facebook', 'youtube', 'youtube_id',\n",
+       "       'bioguide_id', 'thomas_id', 'opensecrets_id', 'lis_id', 'fec_ids',\n",
+       "       'cspan_id', 'govtrack_id', 'votesmart_id', 'ballotpedia_id',\n",
+       "       'washington_post_id', 'icpsr_id', 'wikipedia_id', 'last_name.1',\n",
+       "       'first_name.1', 'middle_name.1', 'suffix.1', 'nickname.1',\n",
+       "       'full_name.1', 'birthday.1', 'gender.1', 'type.1', 'state.1',\n",
+       "       'district.1', 'senate_class.1', 'party.1', 'url.1', 'address.1',\n",
+       "       'phone.1', 'contact_form.1', 'rss_url.1', 'twitter.1', 'facebook.1',\n",
+       "       'youtube.1', 'youtube_id.1', 'bioguide_id.1', 'thomas_id.1',\n",
+       "       'opensecrets_id.1', 'lis_id.1', 'fec_ids.1', 'cspan_id.1',\n",
+       "       'govtrack_id.1', 'votesmart_id.1', 'ballotpedia_id.1',\n",
+       "       'washington_post_id.1', 'icpsr_id.1', 'wikipedia_id.1'],\n",
+       "      dtype='object')"
+      ]
+     },
+     "execution_count": 70,
+     "metadata": {},
+     "output_type": "execute_result"
+    }
+   ],
+   "source": [
+    "#data_df['did'] = data_df.apply(lambda x: generate_did()[0], axis=1)\n",
+    "#data_df['key'] = data_df.apply(lambda x: generate_did()[1], axis=1)\n",
+    "cleaned_df = data_df.copy()\n",
+    "cleaned_df.head()\n",
+    "cleaned_df.isna().sum()\n",
+    "cleaned_df.isna().dropna()\n",
+    "cleaned_df.describe()\n",
+    "cleaned_df.shape\n",
+    "cleaned_df.columns"
+   ]
+  },
+  {
+   "cell_type": "code",
+   "execution_count": 71,
+   "metadata": {},
+   "outputs": [],
+   "source": [
+    "data_did = data_df.copy()\n",
+    "data_did.to_csv(\"data_did.csv\")\n",
+    "data_did"
+   ]
+  },
+  {
+   "cell_type": "code",
+   "execution_count": 76,
+   "metadata": {},
+   "outputs": [
+    {
+     "ename": "ModuleNotFoundError",
+     "evalue": "No module named 'DatasetDict'",
+     "output_type": "error",
+     "traceback": [
+      "\u001b[0;31m---------------------------------------------------------------------------\u001b[0m",
+      "\u001b[0;31mModuleNotFoundError\u001b[0m                       Traceback (most recent call last)",
+      "Cell \u001b[0;32mIn[76], line 2\u001b[0m\n\u001b[1;32m      1\u001b[0m \u001b[38;5;66;03m# Back to dataset\u001b[39;00m\n\u001b[0;32m----> 2\u001b[0m \u001b[38;5;28;01mimport\u001b[39;00m \u001b[38;5;21;01mDatasetDict\u001b[39;00m \n\u001b[1;32m      3\u001b[0m secure_mode \u001b[38;5;241m=\u001b[39m \u001b[38;5;28;01mFalse\u001b[39;00m\n\u001b[1;32m      4\u001b[0m train_split \u001b[38;5;241m=\u001b[39m \u001b[38;5;241m0.8\u001b[39m\n",
+      "\u001b[0;31mModuleNotFoundError\u001b[0m: No module named 'DatasetDict'"
+     ]
+    }
+   ],
+   "source": [
+    "# Back to dataset\n",
+    "\n",
+    "secure_mode = False\n",
+    "train_split = 0.8\n",
+    "test_every = 5\n",
+    "batch_size = 800\n",
+    "\n",
+    "ds = data_did\n",
+    "train_len = int(train_split * len(ds))\n",
+    "test_len = len(ds) - train_len\n",
+    "\n",
+    "print(f\"{train_len} samples for training, {test_len} for testing\")\n",
+    "\n",
+    "train_ds, test_ds = torch.utils.data.random_split(ds, [train_len, test_len])\n"
+   ]
+  },
+  {
+   "cell_type": "code",
+   "execution_count": 78,
+   "metadata": {},
+   "outputs": [
+    {
+     "ename": "TypeError",
+     "evalue": "expected str, bytes or os.PathLike object, not Subset",
+     "output_type": "error",
+     "traceback": [
+      "\u001b[0;31m---------------------------------------------------------------------------\u001b[0m",
+      "\u001b[0;31mTypeError\u001b[0m                                 Traceback (most recent call last)",
+      "Cell \u001b[0;32mIn[78], line 2\u001b[0m\n\u001b[1;32m      1\u001b[0m \u001b[38;5;28;01mfrom\u001b[39;00m \u001b[38;5;21;01mdatasets\u001b[39;00m \u001b[38;5;28;01mimport\u001b[39;00m load_dataset\n\u001b[0;32m----> 2\u001b[0m ds \u001b[38;5;241m=\u001b[39m \u001b[43mload_dataset\u001b[49m\u001b[43m(\u001b[49m\u001b[43mtrain_ds\u001b[49m\u001b[43m,\u001b[49m\u001b[43m \u001b[49m\u001b[43mtest_ds\u001b[49m\u001b[43m,\u001b[49m\u001b[43m \u001b[49m\u001b[43msplit\u001b[49m\u001b[38;5;241;43m=\u001b[39;49m\u001b[43m[\u001b[49m\u001b[38;5;124;43m\"\u001b[39;49m\u001b[38;5;124;43mtrain\u001b[39;49m\u001b[38;5;124;43m\"\u001b[39;49m\u001b[43m,\u001b[49m\u001b[43m \u001b[49m\u001b[38;5;124;43m\"\u001b[39;49m\u001b[38;5;124;43mtest\u001b[39;49m\u001b[38;5;124;43m\"\u001b[39;49m\u001b[43m]\u001b[49m\u001b[43m)\u001b[49m\n",
+      "File \u001b[0;32m~/Library/Caches/pypoetry/virtualenvs/verida-differential-privacy-OB45ac0m-py3.11/lib/python3.11/site-packages/datasets/load.py:2588\u001b[0m, in \u001b[0;36mload_dataset\u001b[0;34m(path, name, data_dir, data_files, split, cache_dir, features, download_config, download_mode, verification_mode, ignore_verifications, keep_in_memory, save_infos, revision, token, use_auth_token, task, streaming, num_proc, storage_options, trust_remote_code, **config_kwargs)\u001b[0m\n\u001b[1;32m   2586\u001b[0m \u001b[38;5;28;01mif\u001b[39;00m data_files \u001b[38;5;129;01mis\u001b[39;00m \u001b[38;5;129;01mnot\u001b[39;00m \u001b[38;5;28;01mNone\u001b[39;00m \u001b[38;5;129;01mand\u001b[39;00m \u001b[38;5;129;01mnot\u001b[39;00m data_files:\n\u001b[1;32m   2587\u001b[0m     \u001b[38;5;28;01mraise\u001b[39;00m \u001b[38;5;167;01mValueError\u001b[39;00m(\u001b[38;5;124mf\u001b[39m\u001b[38;5;124m\"\u001b[39m\u001b[38;5;124mEmpty \u001b[39m\u001b[38;5;124m'\u001b[39m\u001b[38;5;124mdata_files\u001b[39m\u001b[38;5;124m'\u001b[39m\u001b[38;5;124m: \u001b[39m\u001b[38;5;124m'\u001b[39m\u001b[38;5;132;01m{\u001b[39;00mdata_files\u001b[38;5;132;01m}\u001b[39;00m\u001b[38;5;124m'\u001b[39m\u001b[38;5;124m. It should be either non-empty or None (default).\u001b[39m\u001b[38;5;124m\"\u001b[39m)\n\u001b[0;32m-> 2588\u001b[0m \u001b[38;5;28;01mif\u001b[39;00m \u001b[43mPath\u001b[49m\u001b[43m(\u001b[49m\u001b[43mpath\u001b[49m\u001b[43m,\u001b[49m\u001b[43m \u001b[49m\u001b[43mconfig\u001b[49m\u001b[38;5;241;43m.\u001b[39;49m\u001b[43mDATASET_STATE_JSON_FILENAME\u001b[49m\u001b[43m)\u001b[49m\u001b[38;5;241m.\u001b[39mexists():\n\u001b[1;32m   2589\u001b[0m     \u001b[38;5;28;01mraise\u001b[39;00m \u001b[38;5;167;01mValueError\u001b[39;00m(\n\u001b[1;32m   2590\u001b[0m         \u001b[38;5;124m\"\u001b[39m\u001b[38;5;124mYou are trying to load a dataset that was saved using `save_to_disk`. \u001b[39m\u001b[38;5;124m\"\u001b[39m\n\u001b[1;32m   2591\u001b[0m         \u001b[38;5;124m\"\u001b[39m\u001b[38;5;124mPlease use `load_from_disk` instead.\u001b[39m\u001b[38;5;124m\"\u001b[39m\n\u001b[1;32m   2592\u001b[0m     )\n\u001b[1;32m   2594\u001b[0m \u001b[38;5;28;01mif\u001b[39;00m streaming \u001b[38;5;129;01mand\u001b[39;00m num_proc \u001b[38;5;129;01mis\u001b[39;00m \u001b[38;5;129;01mnot\u001b[39;00m \u001b[38;5;28;01mNone\u001b[39;00m:\n",
+      "File \u001b[0;32m/opt/homebrew/Cellar/[email protected]/3.11.9_1/Frameworks/Python.framework/Versions/3.11/lib/python3.11/pathlib.py:871\u001b[0m, in \u001b[0;36mPath.__new__\u001b[0;34m(cls, *args, **kwargs)\u001b[0m\n\u001b[1;32m    869\u001b[0m \u001b[38;5;28;01mif\u001b[39;00m \u001b[38;5;28mcls\u001b[39m \u001b[38;5;129;01mis\u001b[39;00m Path:\n\u001b[1;32m    870\u001b[0m     \u001b[38;5;28mcls\u001b[39m \u001b[38;5;241m=\u001b[39m WindowsPath \u001b[38;5;28;01mif\u001b[39;00m os\u001b[38;5;241m.\u001b[39mname \u001b[38;5;241m==\u001b[39m \u001b[38;5;124m'\u001b[39m\u001b[38;5;124mnt\u001b[39m\u001b[38;5;124m'\u001b[39m \u001b[38;5;28;01melse\u001b[39;00m PosixPath\n\u001b[0;32m--> 871\u001b[0m \u001b[38;5;28mself\u001b[39m \u001b[38;5;241m=\u001b[39m \u001b[38;5;28;43mcls\u001b[39;49m\u001b[38;5;241;43m.\u001b[39;49m\u001b[43m_from_parts\u001b[49m\u001b[43m(\u001b[49m\u001b[43margs\u001b[49m\u001b[43m)\u001b[49m\n\u001b[1;32m    872\u001b[0m \u001b[38;5;28;01mif\u001b[39;00m \u001b[38;5;129;01mnot\u001b[39;00m \u001b[38;5;28mself\u001b[39m\u001b[38;5;241m.\u001b[39m_flavour\u001b[38;5;241m.\u001b[39mis_supported:\n\u001b[1;32m    873\u001b[0m     \u001b[38;5;28;01mraise\u001b[39;00m \u001b[38;5;167;01mNotImplementedError\u001b[39;00m(\u001b[38;5;124m\"\u001b[39m\u001b[38;5;124mcannot instantiate \u001b[39m\u001b[38;5;132;01m%r\u001b[39;00m\u001b[38;5;124m on your system\u001b[39m\u001b[38;5;124m\"\u001b[39m\n\u001b[1;32m    874\u001b[0m                               \u001b[38;5;241m%\u001b[39m (\u001b[38;5;28mcls\u001b[39m\u001b[38;5;241m.\u001b[39m\u001b[38;5;18m__name__\u001b[39m,))\n",
+      "File \u001b[0;32m/opt/homebrew/Cellar/[email protected]/3.11.9_1/Frameworks/Python.framework/Versions/3.11/lib/python3.11/pathlib.py:509\u001b[0m, in \u001b[0;36mPurePath._from_parts\u001b[0;34m(cls, args)\u001b[0m\n\u001b[1;32m    504\u001b[0m \u001b[38;5;129m@classmethod\u001b[39m\n\u001b[1;32m    505\u001b[0m \u001b[38;5;28;01mdef\u001b[39;00m \u001b[38;5;21m_from_parts\u001b[39m(\u001b[38;5;28mcls\u001b[39m, args):\n\u001b[1;32m    506\u001b[0m     \u001b[38;5;66;03m# We need to call _parse_args on the instance, so as to get the\u001b[39;00m\n\u001b[1;32m    507\u001b[0m     \u001b[38;5;66;03m# right flavour.\u001b[39;00m\n\u001b[1;32m    508\u001b[0m     \u001b[38;5;28mself\u001b[39m \u001b[38;5;241m=\u001b[39m \u001b[38;5;28mobject\u001b[39m\u001b[38;5;241m.\u001b[39m\u001b[38;5;21m__new__\u001b[39m(\u001b[38;5;28mcls\u001b[39m)\n\u001b[0;32m--> 509\u001b[0m     drv, root, parts \u001b[38;5;241m=\u001b[39m \u001b[38;5;28;43mself\u001b[39;49m\u001b[38;5;241;43m.\u001b[39;49m\u001b[43m_parse_args\u001b[49m\u001b[43m(\u001b[49m\u001b[43margs\u001b[49m\u001b[43m)\u001b[49m\n\u001b[1;32m    510\u001b[0m     \u001b[38;5;28mself\u001b[39m\u001b[38;5;241m.\u001b[39m_drv \u001b[38;5;241m=\u001b[39m drv\n\u001b[1;32m    511\u001b[0m     \u001b[38;5;28mself\u001b[39m\u001b[38;5;241m.\u001b[39m_root \u001b[38;5;241m=\u001b[39m root\n",
+      "File \u001b[0;32m/opt/homebrew/Cellar/[email protected]/3.11.9_1/Frameworks/Python.framework/Versions/3.11/lib/python3.11/pathlib.py:493\u001b[0m, in \u001b[0;36mPurePath._parse_args\u001b[0;34m(cls, args)\u001b[0m\n\u001b[1;32m    491\u001b[0m     parts \u001b[38;5;241m+\u001b[39m\u001b[38;5;241m=\u001b[39m a\u001b[38;5;241m.\u001b[39m_parts\n\u001b[1;32m    492\u001b[0m \u001b[38;5;28;01melse\u001b[39;00m:\n\u001b[0;32m--> 493\u001b[0m     a \u001b[38;5;241m=\u001b[39m os\u001b[38;5;241m.\u001b[39mfspath(a)\n\u001b[1;32m    494\u001b[0m     \u001b[38;5;28;01mif\u001b[39;00m \u001b[38;5;28misinstance\u001b[39m(a, \u001b[38;5;28mstr\u001b[39m):\n\u001b[1;32m    495\u001b[0m         \u001b[38;5;66;03m# Force-cast str subclasses to str (issue #21127)\u001b[39;00m\n\u001b[1;32m    496\u001b[0m         parts\u001b[38;5;241m.\u001b[39mappend(\u001b[38;5;28mstr\u001b[39m(a))\n",
+      "\u001b[0;31mTypeError\u001b[0m: expected str, bytes or os.PathLike object, not Subset"
+     ]
+    }
+   ],
+   "source": [
+    "from datasets import load_dataset\n",
+    "ds = load_dataset(train_ds, test_ds, split=[\"train\", \"test\"])\n"
+   ]
+  },
+  {
+   "cell_type": "code",
+   "execution_count": 79,
+   "metadata": {},
+   "outputs": [],
+   "source": [
+    "import torch\n",
+    "import torch.nn as nn\n",
+    "\n",
+    "class CharByteEncoder(nn.Module):\n",
+    "    \"\"\"\n",
+    "    This encoder takes a UTF-8 string and encodes its bytes into a Tensor. It can also\n",
+    "    perform the opposite operation to check a result.\n",
+    "    Examples:\n",
+    "    >>> encoder = CharByteEncoder()\n",
+    "    >>> t = encoder('Ślusàrski')  # returns tensor([256, 197, 154, 108, 117, 115, 195, 160, 114, 115, 107, 105, 257])\n",
+    "    >>> encoder.decode(t)  # returns \"<s>Ślusàrski</s>\"\n",
+    "    \"\"\"\n",
+    "\n",
+    "    def __init__(self):\n",
+    "        super().__init__()\n",
+    "        self.start_token = \"<s>\"\n",
+    "        self.end_token = \"</s>\"\n",
+    "        self.pad_token = \"<pad>\"\n",
+    "\n",
+    "        self.start_idx = 256\n",
+    "        self.end_idx = 257\n",
+    "        self.pad_idx = 258\n",
+    "\n",
+    "    def forward(self, s: str, pad_to=0) -> torch.LongTensor:\n",
+    "        \"\"\"\n",
+    "        Encodes a string. It will append a start token <s> (id=self.start_idx) and an end token </s>\n",
+    "        (id=self.end_idx).\n",
+    "        Args:\n",
+    "            s: The string to encode.\n",
+    "            pad_to: If not zero, pad by appending self.pad_idx until string is of length `pad_to`.\n",
+    "                Defaults to 0.\n",
+    "        Returns:\n",
+    "            The encoded LongTensor of indices.\n",
+    "        \"\"\"\n",
+    "        encoded = s.encode()\n",
+    "        n_pad = pad_to - len(encoded) if pad_to > len(encoded) else 0\n",
+    "        return torch.LongTensor(\n",
+    "            [self.start_idx]\n",
+    "            + [c for c in encoded]  # noqa\n",
+    "            + [self.end_idx]\n",
+    "            + [self.pad_idx for _ in range(n_pad)]\n",
+    "        )\n",
+    "\n",
+    "    def decode(self, char_ids_tensor: torch.LongTensor) -> str:\n",
+    "        \"\"\"\n",
+    "        The inverse of `forward`. Keeps the start, end, and pad indices.\n",
+    "        \"\"\"\n",
+    "        char_ids = char_ids_tensor.cpu().detach().tolist()\n",
+    "\n",
+    "        out = []\n",
+    "        buf = []\n",
+    "        for c in char_ids:\n",
+    "            if c < 256:\n",
+    "                buf.append(c)\n",
+    "            else:\n",
+    "                if buf:\n",
+    "                    out.append(bytes(buf).decode())\n",
+    "                    buf = []\n",
+    "                if c == self.start_idx:\n",
+    "                    out.append(self.start_token)\n",
+    "                elif c == self.end_idx:\n",
+    "                    out.append(self.end_token)\n",
+    "                elif c == self.pad_idx:\n",
+    "                    out.append(self.pad_token)\n",
+    "\n",
+    "        if buf:  # in case some are left\n",
+    "            out.append(bytes(buf).decode())\n",
+    "        return \"\".join(out)\n",
+    "\n",
+    "    def __len__(self):\n",
+    "        \"\"\"\n",
+    "        The length of our encoder space. This is fixed to 256 (one byte) + 3 special chars\n",
+    "        (start, end, pad).\n",
+    "        Returns:\n",
+    "            259\n",
+    "        \"\"\"\n",
+    "        return 259"
+   ]
+  },
+  {
+   "cell_type": "code",
+   "execution_count": 80,
+   "metadata": {},
+   "outputs": [],
+   "source": [
+    "from torch.nn.utils.rnn import pad_sequence\n",
+    "\n",
+    "def padded_collate(batch, padding_idx=0):\n",
+    "    x = pad_sequence(\n",
+    "        [elem[0] for elem in batch], batch_first=True, padding_value=padding_idx\n",
+    "    )\n",
+    "    y = torch.stack([elem[1] for elem in batch]).long()\n",
+    "\n",
+    "    return x, y"
+   ]
+  },
+  {
+   "cell_type": "code",
+   "execution_count": 74,
+   "metadata": {},
+   "outputs": [],
+   "source": [
+    "from torch.utils.data import Dataset\n",
+    "from pathlib import Path\n",
+    "\n",
+    "\n",
+    "class NamesDataset(Dataset):\n",
+    "    def __init__(self, root):\n",
+    "        self.root = Path(root)\n",
+    "\n",
+    "        self.labels = list({langfile.stem for langfile in self.root.iterdir()})\n",
+    "        self.labels_dict = {label: i for i, label in enumerate(self.labels)}\n",
+    "        self.encoder = CharByteEncoder()\n",
+    "        self.samples = self.construct_samples()\n",
+    "\n",
+    "    def __getitem__(self, i):\n",
+    "        return self.samples[i]\n",
+    "\n",
+    "    def __len__(self):\n",
+    "        return len(self.samples)\n",
+    "\n",
+    "    def construct_samples(self):\n",
+    "        samples = []\n",
+    "        for langfile in self.root.iterdir():\n",
+    "            label_name = langfile.stem\n",
+    "            label_id = self.labels_dict[label_name]\n",
+    "            with open(langfile, \"r\") as fin:\n",
+    "                for row in fin:\n",
+    "                    samples.append(\n",
+    "                        (self.encoder(row.strip()), torch.tensor(label_id).long())\n",
+    "                    )\n",
+    "        return samples\n",
+    "\n",
+    "    def label_count(self):\n",
+    "        cnt = Counter()\n",
+    "        for _x, y in self.samples:\n",
+    "            label = self.labels[int(y)]\n",
+    "            cnt[label] += 1\n",
+    "        return cnt\n",
+    "\n",
+    "\n",
+    "VOCAB_SIZE = 256 + 3  # 256 alternatives in one byte, plus 3 special characters."
+   ]
+  },
+  {
+   "cell_type": "code",
+   "execution_count": 81,
+   "metadata": {},
+   "outputs": [],
+   "source": [
+    "# Data Loaders\n",
+    "from torch.utils.data import DataLoader\n",
+    "\n",
+    "train_loader = DataLoader(\n",
+    "    train_ds,\n",
+    "    batch_size=batch_size,\n",
+    "    pin_memory=True,\n",
+    "    collate_fn=padded_collate,\n",
+    ")\n",
+    "\n",
+    "test_loader = DataLoader(\n",
+    "    test_ds,\n",
+    "    batch_size=2 * batch_size,\n",
+    "    shuffle=False,\n",
+    "    pin_memory=True,\n",
+    "    collate_fn=padded_collate,\n",
+    ")"
+   ]
+  },
+  {
+   "cell_type": "code",
+   "execution_count": 85,
+   "metadata": {},
+   "outputs": [],
+   "source": [
+    "import pandas as pd\n",
+    "\n",
+    "df = pd.read_csv(\"hf://datasets/synavate/medical_records_did/data_did.csv\")"
+   ]
+  },
+  {
+   "cell_type": "code",
+   "execution_count": 93,
+   "metadata": {},
+   "outputs": [
+    {
+     "data": {
+      "text/plain": [
+       "Index(['last_name', 'first_name', 'full_name', 'birthday', 'gender', 'type',\n",
+       "       'state', 'district', 'senate_class', 'party', 'url', 'address', 'phone',\n",
+       "       'contact_form', 'rss_url', 'twitter', 'facebook', 'youtube',\n",
+       "       'youtube_id', 'bioguide_id', 'thomas_id', 'opensecrets_id', 'lis_id',\n",
+       "       'fec_ids', 'cspan_id', 'govtrack_id', 'votesmart_id', 'ballotpedia_id',\n",
+       "       'washington_post_id', 'icpsr_id', 'wikipedia_id', 'last_name.1',\n",
+       "       'first_name.1', 'middle_name.1', 'suffix.1', 'nickname.1',\n",
+       "       'full_name.1', 'birthday.1', 'gender.1', 'type.1', 'state.1',\n",
+       "       'district.1', 'senate_class.1', 'party.1', 'url.1', 'address.1',\n",
+       "       'phone.1', 'contact_form.1', 'rss_url.1', 'twitter.1', 'facebook.1',\n",
+       "       'youtube.1', 'youtube_id.1', 'bioguide_id.1', 'thomas_id.1',\n",
+       "       'opensecrets_id.1', 'lis_id.1', 'fec_ids.1', 'cspan_id.1',\n",
+       "       'govtrack_id.1', 'votesmart_id.1', 'ballotpedia_id.1',\n",
+       "       'washington_post_id.1', 'icpsr_id.1', 'wikipedia_id.1'],\n",
+       "      dtype='object')"
+      ]
+     },
+     "execution_count": 93,
+     "metadata": {},
+     "output_type": "execute_result"
+    }
+   ],
+   "source": [
+    "df_drop=df.copy()\n",
+    "df_drop.isnull().drop(index=1)\n",
+    "df_drop.isna().sum()\n",
+    "df_drop.drop(columns=['Unnamed: 0', 'middle_name', 'suffix', 'nickname'], inplace=True)\n",
+    "df_drop.columns"
+   ]
+  },
+  {
+   "cell_type": "code",
+   "execution_count": null,
+   "metadata": {},
+   "outputs": [],
+   "source": []
+  },
+  {
+   "cell_type": "code",
+   "execution_count": null,
+   "metadata": {},
+   "outputs": [],
+   "source": [
+    "!pip install scikit-learn\n",
+    "from sklearn import train_test_split\n",
+    "\n"
+   ]
+  }
+ ],
+ "metadata": {
+  "kernelspec": {
+   "display_name": "Python 3 (ipykernel)",
+   "language": "python",
+   "name": "python3"
+  },
+  "language_info": {
+   "codemirror_mode": {
+    "name": "ipython",
+    "version": 3
+   },
+   "file_extension": ".py",
+   "mimetype": "text/x-python",
+   "name": "python",
+   "nbconvert_exporter": "python",
+   "pygments_lexer": "ipython3",
+   "version": "3.11.9"
+  }
+ },
+ "nbformat": 4,
+ "nbformat_minor": 2
+}

gretel_ai/gretel_exp.md

@@ -0,0 +1,4 @@
+# Using Gretel AI to Redact PII 
+### Applying Gretel.ai's dfferential AI to the equivalent dataset
+
+

medical_records_did

@@ -0,0 +1 @@
+Subproject commit 7d84b291818a39727cd0e0dfe1ccad2c4bbda2dd

pyproject.toml

@@ -0,0 +1,30 @@
+[tool.poetry]
+name = "verida-differential-privacy"
+version = "0.1.0"
+description = "A demonstration of using the PyTorch differential privacy library and a login DiD"
+authors = ["snyata <[email protected]>"]
+license = "MIT"
+readme = "README.md"
+
+[tool.poetry.dependencies]
+python = "^3.10"
+pydantic = "^2.8.2"
+wandb = "^0.17.7"
+python-dotenv = "^1.0.1"
+ipykernel = "^6.29.5"
+datasets = "^2.21.0"
+didkit = "^0.3.3"
+scikit-learn = "^1.5.1"
+matplotlib = "^3.9.2"
+seaborn = "^0.13.2"
+
+[tool.poetry.group.dev.dependencies]
+pytest = "8.3.2"
+pip-tools = "^7.4.1"
+isort = "^5.13.2"
+datasets = "^2.21.0"
+jupyter = "^1.0.0"
+
+[build-system]
+requires = ["poetry-core"]
+build-backend = "poetry.core.masonry.api"

transformer_models/unsloth_model_colab.ipynb

@@ -0,0 +1,466 @@
+{
+ "cells": [
+  {
+   "cell_type": "markdown",
+   "metadata": {},
+   "source": [
+    "# PyTorch Differential Privacy Experiment"
+   ]
+  },
+  {
+   "cell_type": "code",
+   "execution_count": null,
+   "metadata": {},
+   "outputs": [],
+   "source": [
+    "%pip install -qqq torch torchvision opacus numpy pandas\n",
+    "%pip install -qqq wandb datasets tqdm\n"
+   ]
+  },
+  {
+   "cell_type": "code",
+   "execution_count": null,
+   "metadata": {},
+   "outputs": [],
+   "source": [
+    "import os\n",
+    "import torch \n",
+    "from dotenv import load_dotenv\n",
+    "import wandb \n",
+    "import logging\n",
+    "import shutil\n",
+    "import sys\n",
+    "from datetime import datetime, timedelta\n",
+    "\n",
+    "import argparse\n",
+    "from collections import Counter\n",
+    "from pathlib import Path\n",
+    "from statistics import mean\n",
+    "\n",
+    "import torch\n",
+    "import torch.nn as nn\n",
+    "from opacus import PrivacyEngine\n",
+    "from opacus.layers import DPGRU, DPLSTM, DPRNN\n",
+    "from torch.nn.utils.rnn import pad_sequence\n",
+    "from torch.utils.data import DataLoader, Dataset\n",
+    "from tqdm import tqdm, tqdm_notebook"
+   ]
+  },
+  {
+   "cell_type": "code",
+   "execution_count": null,
+   "metadata": {},
+   "outputs": [],
+   "source": [
+    "device = \"mps\" if torch.backends.mps.is_available() else \"cpu\"\n",
+    "if os.path.exists('.env'):\n",
+    "    load_dotenv('.env')\n",
+    "device\n"
+   ]
+  },
+  {
+   "cell_type": "code",
+   "execution_count": null,
+   "metadata": {},
+   "outputs": [],
+   "source": [
+    "logging.basicConfig(\n",
+    "    format=\"%(asctime)s:%(levelname)s:%(message)s\",\n",
+    "    datefmt=\"%m/%d/%Y %H:%M:%S\",\n",
+    "    stream=sys.stdout,\n",
+    ")\n",
+    "logger = logging.getLogger(\"ddp\")\n",
+    "logger.setLevel(level=logging.INFO)\n"
+   ]
+  },
+  {
+   "cell_type": "code",
+   "execution_count": null,
+   "metadata": {},
+   "outputs": [],
+   "source": [
+    "wandb.login(key=os.getenv('WANDB_API_KEY'))\n",
+    "wandb.init(project=\"verida-pii\", name=\"deberta_finetune\")"
+   ]
+  },
+  {
+   "cell_type": "markdown",
+   "metadata": {},
+   "source": [
+    "# Fine Tuning w/ Unsloth (Colab Only)"
+   ]
+  },
+  {
+   "cell_type": "code",
+   "execution_count": null,
+   "metadata": {},
+   "outputs": [],
+   "source": [
+    "# Datasets\n",
+    "# Original data_name = 'Ezi/medical_and_legislators_synthetic'\n",
+    "# Tutorial: https://huggingface.co/blog/Andyrasika/finetune-unsloth-qlora\n"
+   ]
+  },
+  {
+   "cell_type": "code",
+   "execution_count": null,
+   "metadata": {},
+   "outputs": [],
+   "source": [
+    "# Get the major and minor version of the current CUDA device (GPU)\n",
+    "major_version, minor_version = torch.cuda.get_device_capability()\n",
+    "\n",
+    "# Apply the following if the GPU has Ampere or Hopper architecture (RTX 30xx, RTX 40xx, A100, H100, L40, etc.)\n",
+    "if major_version >= 8:\n",
+    "    # Install the Unsloth library for Ampere and Hopper architecture from GitHub\n",
+    "    !pip install \"unsloth[colab_ampere] @ git+https://github.com/unslothai/unsloth.git\" -q\n",
+    "\n",
+    "# Apply the following for older GPUs (V100, Tesla T4, RTX 20xx, etc.)\n",
+    "else:\n",
+    "    # Install the Unsloth library for older GPUs from GitHub\n",
+    "    !pip install \"unsloth[colab] @ git+https://github.com/unslothai/unsloth.git\" -q\n",
+    "\n",
+    "# Placeholder statement (does nothing)\n",
+    "pass\n",
+    "\n",
+    "# Install the Hugging Face Transformers library from GitHub, which allows native 4-bit loading\n",
+    "!pip install \"git+https://github.com/huggingface/transformers.git\" -q\n",
+    "\n"
+   ]
+  },
+  {
+   "cell_type": "code",
+   "execution_count": null,
+   "metadata": {},
+   "outputs": [],
+   "source": [
+    "from transformers import AutoTokenizer, AutoModelForTokenClassification\n",
+    "\n",
+    "tokenizer = AutoTokenizer.from_pretrained(\"lakshyakh93/deberta_finetuned_pii\")\n",
+    "model = AutoModelForTokenClassification.from_pretrained(\"lakshyakh93/deberta_finetuned_pii\")"
+   ]
+  },
+  {
+   "cell_type": "code",
+   "execution_count": null,
+   "metadata": {},
+   "outputs": [],
+   "source": [
+    "model = FastLanguageModel.get_peft_model(\n",
+    "    model,\n",
+    "    # Specify the existing model\n",
+    "\n",
+    "    r=16,  # Choose any positive number! Recommended values include 8, 16, 32, 64, 128, etc.\n",
+    "    # Rank parameter for LoRA. The smaller this value, the fewer parameters will be modified.\n",
+    "\n",
+    "    target_modules=[\"q_proj\", \"k_proj\", \"v_proj\", \"o_proj\",\n",
+    "                    \"gate_proj\", \"up_proj\", \"down_proj\",],\n",
+    "    # Specify the modules to which LoRA will be applied\n",
+    "\n",
+    "    lora_alpha=16,\n",
+    "    # Alpha parameter for LoRA. This value determines the strength of the applied LoRA.\n",
+    "\n",
+    "    lora_dropout=0,  # Currently, only supports dropout = 0\n",
+    "    # Dropout rate for LoRA. Currently supports only 0.\n",
+    "\n",
+    "    bias=\"none\",  # Currently, only supports bias = \"none\"\n",
+    "    # Bias usage setting. Currently supports only the setting without bias.\n",
+    "\n",
+    "    use_gradient_checkpointing=True,\n",
+    "    # Whether to use gradient checkpointing to improve memory efficiency\n",
+    "\n",
+    "    random_state=3407,\n",
+    "    # Seed value for random number generation\n",
+    "\n",
+    "    max_seq_length=max_seq_length,\n",
+    "    # Set the maximum sequence length\n",
+    ")"
+   ]
+  },
+  {
+   "cell_type": "code",
+   "execution_count": null,
+   "metadata": {},
+   "outputs": [],
+   "source": [
+    "# @TODO - Add the relevant prompt\n",
+    "alpaca_prompt = \"\"\"Below is an instruction that describes a task, paired with an input that provides further context. Write a response that appropriately completes the request.\n",
+    "\n",
+    "### Instruction:\n",
+    "{}\n",
+    "\n",
+    "### Input:\n",
+    "{}\n",
+    "\n",
+    "### Response:\n",
+    "{}\"\"\"\n",
+    "# Define the prompt format for the Alpaca dataset\n",
+    "\n",
+    "def formatting_prompts_func(examples):\n",
+    "    # Define a function to format each example in the dataset\n",
+    "\n",
+    "    instructions = examples[\"instruction\"]\n",
+    "    inputs       = examples[\"input\"]\n",
+    "    outputs      = examples[\"output\"]\n",
+    "    # Get instructions, inputs, and outputs\n",
+    "\n",
+    "    texts = []\n",
+    "    for instruction, input, output in zip(instructions, inputs, outputs):\n",
+    "        # Generate text by combining instructions, inputs, and outputs\n",
+    "\n",
+    "        text = alpaca_prompt.format(instruction, input, output)\n",
+    "        # Format the text according to the prompt format\n",
+    "\n",
+    "        texts.append(text)\n",
+    "    return { \"text\" : texts, }\n",
+    "    # Return a list of formatted texts\n",
+    "\n",
+    "pass\n",
+    "# Placeholder (does nothing)\n",
+    "\n",
+    "from datasets import load_dataset\n",
+    "# Import the load_dataset function from the datasets library\n",
+    "\n",
+    "dataset = load_dataset(\"yahma/alpaca-cleaned\", split=\"train\")\n",
+    "# Load the training data of the cleaned version of the Alpaca dataset from yahma\n",
+    "\n",
+    "dataset = dataset.map(formatting_prompts_func, batched=True,)"
+   ]
+  },
+  {
+   "cell_type": "code",
+   "execution_count": null,
+   "metadata": {},
+   "outputs": [],
+   "source": [
+    "from trl import SFTTrainer\n",
+    "# Import SFTTrainer from the TRL library\n",
+    "\n",
+    "from transformers import TrainingArguments\n",
+    "# Import TrainingArguments from the Transformers library\n",
+    "\n",
+    "trainer = SFTTrainer(\n",
+    "    # Initialize the SFTTrainer\n",
+    "\n",
+    "    model=model,\n",
+    "    # Specify the model to be used\n",
+    "\n",
+    "    train_dataset=dataset,\n",
+    "    # Specify the training dataset\n",
+    "\n",
+    "    dataset_text_field=\"text\",\n",
+    "    # Specify the text field in the dataset\n",
+    "\n",
+    "    max_seq_length=max_seq_length,\n",
+    "    # Specify the maximum sequence length\n",
+    "\n",
+    "    args=TrainingArguments(\n",
+    "        # Specify training arguments\n",
+    "\n",
+    "        per_device_train_batch_size=2,\n",
+    "        # Specify the training batch size per device\n",
+    "\n",
+    "        gradient_accumulation_steps=4,\n",
+    "        # Specify the number of steps for gradient accumulation\n",
+    "\n",
+    "        warmup_steps=5,\n",
+    "        # Specify the number of warm-up steps\n",
+    "\n",
+    "        max_steps=20,\n",
+    "        # Specify the maximum number of steps\n",
+    "\n",
+    "        learning_rate=2e-4,\n",
+    "        # Specify the learning rate\n",
+    "\n",
+    "        fp16=not torch.cuda.is_bf16_supported(),\n",
+    "        # Set whether to use 16-bit floating-point precision (fp16)\n",
+    "\n",
+    "        bf16=torch.cuda.is_bf16_supported(),\n",
+    "        # Set whether to use Bfloat16\n",
+    "\n",
+    "        logging_steps=1,\n",
+    "        # Specify the logging steps\n",
+    "\n",
+    "        optim=\"adamw_8bit\",\n",
+    "        # Specify the optimizer (here using 8-bit AdamW)\n",
+    "\n",
+    "        weight_decay=0.01,\n",
+    "        # Specify the weight decay value\n",
+    "\n",
+    "        lr_scheduler_type=\"linear\",\n",
+    "        # Specify the type of learning rate scheduler (linear)\n",
+    "\n",
+    "        seed=3407,\n",
+    "        # Specify the random seed\n",
+    "\n",
+    "        output_dir=\"outputs\",\n",
+    "        # Specify the output directory\n",
+    "\n",
+    "    ),\n",
+    ")"
+   ]
+  },
+  {
+   "cell_type": "code",
+   "execution_count": null,
+   "metadata": {},
+   "outputs": [],
+   "source": [
+    "gpu_stats = torch.cuda.get_device_properties(0)\n",
+    "# Get properties of the GPU device at index 0\n",
+    "\n",
+    "start_gpu_memory = round(torch.cuda.max_memory_reserved() / 1024 / 1024 / 1024, 3)\n",
+    "# Get the maximum reserved GPU memory in GB and round to 3 decimal places\n",
+    "\n",
+    "max_memory = round(gpu_stats.total_memory / 1024 / 1024 / 1024, 3)\n",
+    "# Get the total GPU memory in GB and round to 3 decimal places\n",
+    "\n",
+    "print(f\"GPU = {gpu_stats.name}. Max memory = {max_memory} GB.\")\n",
+    "# Display the GPU name and maximum memory\n",
+    "\n",
+    "print(f\"{start_gpu_memory} GB of memory reserved.\")\n",
+    "# Display the reserved memory amount"
+   ]
+  },
+  {
+   "cell_type": "code",
+   "execution_count": null,
+   "metadata": {},
+   "outputs": [],
+   "source": [
+    "trainer_stats = trainer.train()"
+   ]
+  },
+  {
+   "cell_type": "markdown",
+   "metadata": {},
+   "source": [
+    "# Convert to GGUF"
+   ]
+  },
+  {
+   "cell_type": "code",
+   "execution_count": null,
+   "metadata": {},
+   "outputs": [],
+   "source": [
+    "def colab_quantize_to_gguf(save_directory, quantization_method=\"q4_k_m\"):\n",
+    "    # Define a function for conversion to GGUF\n",
+    "\n",
+    "    from transformers.models.llama.modeling_llama import logger\n",
+    "    import os\n",
+    "    # Import necessary libraries\n",
+    "\n",
+    "    logger.warning_once(\n",
+    "        \"Unsloth: `colab_quantize_to_gguf` is still in development mode.\\n\"\\\n",
+    "        \"If anything errors or breaks, please file a ticket on Github.\\n\"\\\n",
+    "        \"Also, if you used this successfully, please tell us on Discord!\"\n",
+    "    )\n",
+    "    # Warn that it's still in development mode and encourage reporting any issues\n",
+    "\n",
+    "    # From https://mlabonne.github.io/blog/posts/Quantize_Llama_2_models_using_ggml.html\n",
+    "    ALLOWED_QUANTS = \\\n",
+    "    {\n",
+    "        # Define currently allowed quantization methods\n",
+    "        # Including descriptions for each quantization method\n",
+    "        \"q2_k\"   : \"Uses Q4_K for the attention.vw and feed_forward.w2 tensors, Q2_K for the other tensors.\",\n",
+    "        \"q3_k_l\" : \"Uses Q5_K for the attention.wv, attention.wo, and feed_forward.w2 tensors, else Q3_K\",\n",
+    "        \"q3_k_m\" : \"Uses Q4_K for the attention.wv, attention.wo, and feed_forward.w2 tensors, else Q3_K\",\n",
+    "        \"q3_k_s\" : \"Uses Q3_K for all tensors\",\n",
+    "        \"q4_0\"   : \"Original quant method, 4-bit.\",\n",
+    "        \"q4_1\"   : \"Higher accuracy than q4_0 but not as high as q5_0. However has quicker inference than q5 models.\",\n",
+    "        \"q4_k_m\" : \"Uses Q6_K for half of the attention.wv and feed_forward.w2 tensors, else Q4_K\",\n",
+    "        \"q4_k_s\" : \"Uses Q4_K for all tensors\",\n",
+    "        \"q5_0\"   : \"Higher accuracy, higher resource usage and slower inference.\",\n",
+    "        \"q5_1\"   : \"Even higher accuracy, resource usage and slower inference.\",\n",
+    "        \"q5_k_m\" : \"Uses Q6_K for half of the attention.wv and feed_forward.w2 tensors, else Q5_K\",\n",
+    "        \"q5_k_s\" : \"Uses Q5_K for all tensors\",\n",
+    "        \"q6_k\"   : \"Uses Q8_K for all tensors\",\n",
+    "        \"q8_0\"   : \"Almost indistinguishable from float16. High resource use and slow. Not recommended for most users.\",\n",
+    "    }\n",
+    "\n",
+    "    if quantization_method not in ALLOWED_QUANTS.keys():\n",
+    "        # If the specified quantization method is not allowed, raise an error\n",
+    "        error = f\"Unsloth: Quant method = [{quantization_method}] not supported. Choose from below:\\n\"\n",
+    "        for key, value in ALLOWED_QUANTS.items():\n",
+    "            error += f\"[{key}] => {value}\\n\"\n",
+    "        raise RuntimeError(error)\n",
+    "\n",
+    "    # Display information about the conversion\n",
+    "    print_info = \\\n",
+    "        f\"==((====))==  Unsloth: Conversion from QLoRA to GGUF information\\n\"\\\n",
+    "        f\"   \\\\\\   /|    [0] Installing llama.cpp will take 3 minutes.\\n\"\\\n",
+    "        f\"O^O/ \\_/ \\\\    [1] Converting HF to GUUF 16bits will take 3 minutes.\\n\"\\\n",
+    "        f\"\\        /    [2] Converting GGUF 16bits to q4_k_m will take 20 minutes.\\n\"\\\n",
+    "        f' \"-____-\"     In total, you will have to wait around 26 minutes.\\n'\n",
+    "    print(print_info)\n",
+    "    # Display information about the conversion process\n",
+    "\n",
+    "    if not os.path.exists(\"llama.cpp\"):\n",
+    "        # If llama.cpp does not exist, install it\n",
+    "        print(\"Unsloth: [0] Installing llama.cpp. This will take 3 minutes...\")\n",
+    "        !git clone https://github.com/ggerganov/llama.cpp\n",
+    "        !cd llama.cpp && make clean && LLAMA_CUBLAS=1 make -j\n",
+    "        !pip install gguf protobuf\n",
+    "        pass\n",
+    "\n",
+    "    print(\"Unsloth: Starting conversion from HF to GGUF 16bit...\")\n",
+    "    # Display that conversion from HF to GGUF 16bit is starting\n",
+    "    # print(\"Unsloth: [1] Converting HF into GGUF 16bit. This will take 3 minutes...\")\n",
+    "    !python llama.cpp/convert.py {save_directory} \\\n",
+    "        --outfile {save_directory}-unsloth.gguf \\\n",
+    "        --outtype f16\n",
+    "\n",
+    "    print(\"Unsloth: Starting conversion from GGUF 16bit to q4_k_m...\")\n",
+    "    # Display that conversion from GGUF 16bit to the specified quantization method is starting\n",
+    "    # print(\"Unsloth: [2] Converting GGUF 16bit into q4_k_m. This will take 20 minutes...\")\n",
+    "    final_location = f\"./{save_directory}-{quantization_method}-unsloth.gguf\"\n",
+    "    !./llama.cpp/quantize ./{save_directory}-unsloth.gguf \\\n",
+    "        {final_location} {quantization_method}\n",
+    "\n",
+    "    print(f\"Unsloth: Output location: {final_location}\")\n",
+    "    # Display the output location of the converted file\n",
+    "pass"
+   ]
+  },
+  {
+   "cell_type": "code",
+   "execution_count": null,
+   "metadata": {},
+   "outputs": [],
+   "source": [
+    "from unsloth import unsloth_save_model\n",
+    "# Import the unsloth_save_model function from the Unsloth library\n",
+    "\n",
+    "# unsloth_save_model has the same args as model.save_pretrained\n",
+    "# unsloth_save_model has the same arguments as model.save_pretrained\n",
+    "unsloth_save_model(model, tokenizer, \"output_model\", push_to_hub=False, token=None)\n",
+    "# Save the model and tokenizer as \"output_model\". Do not push to the Hugging Face Hub\n",
+    "\n",
+    "colab_quantize_to_gguf(\"output_model\", quantization_method=\"q4_k_m\")\n",
+    "# Convert \"output_model\" to GGUF format. Use the quantization method \"q4_k_m\""
+   ]
+  }
+ ],
+ "metadata": {
+  "kernelspec": {
+   "display_name": "Python 3 (ipykernel)",
+   "language": "python",
+   "name": "python3"
+  },
+  "language_info": {
+   "codemirror_mode": {
+    "name": "ipython",
+    "version": 3
+   },
+   "file_extension": ".py",
+   "mimetype": "text/x-python",
+   "name": "python",
+   "nbconvert_exporter": "python",
+   "pygments_lexer": "ipython3",
+   "version": "3.11.9"
+  }
+ },
+ "nbformat": 4,
+ "nbformat_minor": 2
+}