-
Notifications
You must be signed in to change notification settings - Fork 0
/
Copy pathXAI_controller.py
69 lines (48 loc) · 2.25 KB
/
XAI_controller.py
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
import pandas as pd
import shap
import lime
import lime.lime_tabular
import numpy as np
import xai
import traceback
import matplotlib.pyplot as plt
def executeAutoML(function, name, *func_args):
print(f"\n\n\n#################### {name} ####################\n\n\n")
try:
function(*func_args)
except:
traceback.print_exc()
def processCategorical(X):
# Process categories. Lime only accepts categories when processed into int
# https://github.com/marcotcr/lime/issues/346
cat_features = []
cat_names = {}
for i, dtype in enumerate(X.dtypes):
if dtype.name == "category":
# Save index to tell lime where the categorical columns are
cat_features.append(i)
# Process the corresponding column
# Get all unique values in the category column
unique_values = X.iloc[:, i].unique()
# Add the values to the names dictionary required by lime
cat_names.update({i: list(unique_values.categories.values)})
# Replace values in dataframe
X.iloc[:, i] = X.iloc[:, i].map({val: i for i, val in enumerate(unique_values.categories.values)})
if dtype.name == "bool":
cat_features.append(i)
unique_values = [False, True]
cat_names.update({i: unique_values})
X.iloc[:, i] = X.iloc[:, i].map({val: i for i, val in enumerate(unique_values)})
X = X.fillna(0)
return X, cat_names, cat_features
def explainLime(automl, X):
X, cat_names, cat_features = processCategorical(X)
def prediction_probability(X, cols=X.columns, dtypes=X.dtypes):
df = pd.DataFrame(data=X, columns=cols)
df = df.astype(dtype=dict(zip(cols, dtypes.values)))
return automl.predict_proba(df)
explainer = lime.lime_tabular.LimeTabularExplainer(np.array(X), feature_names=X.columns, discretize_continuous=True,
categorical_features=cat_features, categorical_names=cat_names)
i = np.random.randint(0, X.shape[0])
exp = explainer.explain_instance(np.array(X)[i], prediction_probability, num_features=2, top_labels=1)
exp.save_to_file('lime_out.html', labels=None, predict_proba=True, show_predicted_value=True)