File size: 3,410 Bytes
29b9fa1 989839e 29b9fa1 989839e 29b9fa1 989839e 29b9fa1 989839e 29b9fa1 b82e694 989839e |
1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60 61 62 63 64 65 66 67 68 69 70 71 72 73 74 75 76 77 78 |
import json
from itertools import product
import pandas as pd
parameters_min_e_freq = [1, 2, 3, 4]
parameters_max_p_freq = [100, 50, 25, 10]
stats = []
predicate = {}
for min_e_freq, max_p_freq in product(parameters_min_e_freq, parameters_max_p_freq):
with open(f"data/t_rex.filter_unified.min_entity_{min_e_freq}_max_predicate_{max_p_freq}.train.jsonl") as f:
train = [json.loads(i) for i in f.read().split('\n') if len(i) > 0]
df_train = pd.DataFrame(train)
with open(f"data/t_rex.filter_unified.min_entity_{min_e_freq}_max_predicate_{max_p_freq}.validation.jsonl") as f:
validation = [json.loads(i) for i in f.read().split('\n') if len(i) > 0]
df_validation = pd.DataFrame(validation)
with open(f"data/t_rex.filter_unified.min_entity_{min_e_freq}_max_predicate_{max_p_freq}.jsonl") as f:
full = [json.loads(i) for i in f.read().split('\n') if len(i) > 0]
df_full = pd.DataFrame(full)
predicate[f"min_entity_{min_e_freq}_max_predicate_{max_p_freq}"] = df_full['predicate'].unique().tolist()
stats.append({
"data": f"filter_unified.min_entity_{min_e_freq}_max_predicate_{max_p_freq}",
"number of triples (train)": len(train),
"number of triples (validation)": len(validation),
"number of triples (all)": len(full),
"number of unique predicates (train)": len(df_train['predicate'].unique()),
"number of unique predicates (validation)": len(df_validation['predicate'].unique()),
"number of unique predicates (all)": len(df_full['predicate'].unique()),
"number of unique entities (train)": len(
list(set(df_train['object'].unique().tolist() + df_train['subject'].unique().tolist()))),
"number of unique entities (validation)": len(
list(set(df_validation['object'].unique().tolist() + df_validation['subject'].unique().tolist()))),
"number of unique entities (all)": len(
list(set(df_full['object'].unique().tolist() + df_full['subject'].unique().tolist())))
})
df = pd.DataFrame(stats)
df.index = df.pop("data")
for c in df.columns:
df.loc[:, c] = df[c].map('{:,d}'.format)
print(df.to_markdown())
with open(f"data/t_rex.filter_unified.test.jsonl") as f:
test = [json.loads(i) for i in f.read().split('\n') if len(i) > 0]
df_test = pd.DataFrame(test)
predicate["test"] = df_test['predicate'].unique().tolist()
df_test = pd.DataFrame([{
"number of triples (test)": len(df_test),
"number of unique predicates (test)": len(df_test['predicate'].unique()),
"number of unique entities (test)": len(
list(set(df_test['object'].unique().tolist() + df_test['subject'].unique().tolist()))
)
}])
for c in df_test.columns:
df_test.loc[:, c] = df_test[c].map('{:,d}'.format)
print(df_test.to_markdown(index=False))
df["number of triples (test)"] = df_test["number of triples (test)"].values[0]
df["number of unique predicates (test)"] = df_test["number of unique predicates (test)"].values[0]
df["number of unique entities (test)"] = df_test["number of unique entities (test)"].values[0]
df.pop("number of triples (all)")
df.pop("number of unique predicates (all)")
df.pop("number of unique entities (all)")
df = df[sorted(df.columns)]
df.to_csv("data/stats.csv")
# predicates in training vs test
with open("data/predicates.json", "w") as f:
json.dump(predicate, f)
|