chore: update
Browse files
dev.py
CHANGED
|
@@ -1,3 +1,5 @@
|
|
|
|
|
|
|
|
| 1 |
import shutil
|
| 2 |
|
| 3 |
from pathlib import Path
|
|
@@ -7,7 +9,8 @@ import pandas as pd
|
|
| 7 |
from concrete.ml.sklearn import LogisticRegression as ConcreteLogisticRegression
|
| 8 |
from concrete.ml.deployment import FHEModelDev
|
| 9 |
|
| 10 |
-
|
|
|
|
| 11 |
TRAINING_FILE_NAME = "./data/Training_preprocessed.csv"
|
| 12 |
TESTING_FILE_NAME = "./data/Testing_preprocessed.csv"
|
| 13 |
|
|
@@ -15,8 +18,6 @@ TESTING_FILE_NAME = "./data/Testing_preprocessed.csv"
|
|
| 15 |
df_train = pd.read_csv(TRAINING_FILE_NAME)
|
| 16 |
df_test = pd.read_csv(TESTING_FILE_NAME)
|
| 17 |
|
| 18 |
-
print(df_train.shape)
|
| 19 |
-
print(df_train.columns)
|
| 20 |
# Split the data into X_train, y_train, X_test_, y_test sets
|
| 21 |
TARGET_COLUMN = ["prognosis_encoded", "prognosis"]
|
| 22 |
|
|
@@ -26,14 +27,17 @@ y_test = df_test[TARGET_COLUMN[0]].values.flatten()
|
|
| 26 |
X_train = df_train.drop(TARGET_COLUMN, axis=1)
|
| 27 |
X_test = df_test.drop(TARGET_COLUMN, axis=1)
|
| 28 |
|
|
|
|
|
|
|
| 29 |
# Models parameters
|
| 30 |
optimal_param = {"C": 0.9, "n_bits": 13, "solver": "sag", "multi_class": "auto"}
|
| 31 |
|
| 32 |
-
# Concrete ML model
|
| 33 |
clf = ConcreteLogisticRegression(**optimal_param)
|
| 34 |
|
|
|
|
| 35 |
clf.fit(X_train, y_train)
|
| 36 |
|
|
|
|
| 37 |
fhe_circuit = clf.compile(X_train)
|
| 38 |
|
| 39 |
fhe_circuit.client.keygen(force=False)
|
|
|
|
| 1 |
+
"""Generating deployment files."""
|
| 2 |
+
|
| 3 |
import shutil
|
| 4 |
|
| 5 |
from pathlib import Path
|
|
|
|
| 9 |
from concrete.ml.sklearn import LogisticRegression as ConcreteLogisticRegression
|
| 10 |
from concrete.ml.deployment import FHEModelDev
|
| 11 |
|
| 12 |
+
|
| 13 |
+
# Data files location
|
| 14 |
TRAINING_FILE_NAME = "./data/Training_preprocessed.csv"
|
| 15 |
TESTING_FILE_NAME = "./data/Testing_preprocessed.csv"
|
| 16 |
|
|
|
|
| 18 |
df_train = pd.read_csv(TRAINING_FILE_NAME)
|
| 19 |
df_test = pd.read_csv(TESTING_FILE_NAME)
|
| 20 |
|
|
|
|
|
|
|
| 21 |
# Split the data into X_train, y_train, X_test_, y_test sets
|
| 22 |
TARGET_COLUMN = ["prognosis_encoded", "prognosis"]
|
| 23 |
|
|
|
|
| 27 |
X_train = df_train.drop(TARGET_COLUMN, axis=1)
|
| 28 |
X_test = df_test.drop(TARGET_COLUMN, axis=1)
|
| 29 |
|
| 30 |
+
# Concrete ML model
|
| 31 |
+
|
| 32 |
# Models parameters
|
| 33 |
optimal_param = {"C": 0.9, "n_bits": 13, "solver": "sag", "multi_class": "auto"}
|
| 34 |
|
|
|
|
| 35 |
clf = ConcreteLogisticRegression(**optimal_param)
|
| 36 |
|
| 37 |
+
# Fit the model
|
| 38 |
clf.fit(X_train, y_train)
|
| 39 |
|
| 40 |
+
# Compile the model
|
| 41 |
fhe_circuit = clf.compile(X_train)
|
| 42 |
|
| 43 |
fhe_circuit.client.keygen(force=False)
|