def explain(self, doc, truncate_len=512, all_targets=False):
"""
Highlights text to explain prediction
Args:
doc (str): text of documnet
truncate_len(int): truncate document to this many words
all_targets(bool): If True, show visualization for
each target.
"""
try:
import eli5
from eli5.lime import TextExplainer
except:
msg = 'ktrain requires a forked version of eli5 to support tf.keras. '+\
'Install with: pip3 install git+https://github.com/amaiya/eli5@tfkeras_0_10_1'
warnings.warn(msg)
return
prediction = [self.predict(doc)] if not all_targets else None
if not isinstance(doc, str): raise Exception('text must of type str')
if self.preproc.is_nospace_lang():
doc = self.preproc.process_chinese([doc])
doc = doc[0]
doc = ' '.join(doc.split()[:truncate_len])
te = TextExplainer(random_state=42)
_ = te.fit(doc, self.predict_proba)
return te.show_prediction(target_names=self.preproc.get_classes(),
targets=prediction)
def explain(self,
doc,
truncate_len=512,
all_targets=False,
n_samples=2500):
"""
Highlights text to explain prediction
Args:
doc (str): text of documnet
truncate_len(int): truncate document to this many words
all_targets(bool): If True, show visualization for
each target.
n_samples(int): number of samples to generate and train on.
Larger values give better results, but will take more time.
Lower this value if explain is taking too long.
"""
is_array, is_pair = detect_text_format(doc)
if is_pair:
warnings.warn(
"currently_unsupported: explain does not currently support sentence pair classification"
)
return
if not self.c:
warnings.warn(
"currently_unsupported: explain does not support text regression"
)
return
try:
import eli5
from eli5.lime import TextExplainer
except:
msg = (
"ktrain requires a forked version of eli5 to support tf.keras. "
+
"Install with: pip install https://github.com/amaiya/eli5/archive/refs/heads/tfkeras_0_10_1.zip"
)
warnings.warn(msg)
return
if (not hasattr(eli5, "KTRAIN_ELI5_TAG")
or eli5.KTRAIN_ELI5_TAG != KTRAIN_ELI5_TAG):
msg = (
"ktrain requires a forked version of eli5 to support tf.keras. It is either missing or not up-to-date. "
+
"Uninstall the current version and install/re-install the fork with: pip install https://github.com/amaiya/eli5/archive/refs/heads/tfkeras_0_10_1.zip"
)
warnings.warn(msg)
return
if not isinstance(doc, str):
raise TypeError("text must of type str")
prediction = [self.predict(doc)] if not all_targets else None
if self.preproc.is_nospace_lang():
doc = self.preproc.process_chinese([doc])
doc = doc[0]
doc = " ".join(doc.split()[:truncate_len])
te = TextExplainer(random_state=42, n_samples=n_samples)
_ = te.fit(doc, self.predict_proba)
return te.show_prediction(target_names=self.preproc.get_classes(),
targets=prediction)
def eli5visual(pData, pDesc, Idx, pAccountName, pVec, nTopKeywrd, pRootDir):
try:
for i in range(len(Idx)):
if Idx[i] <= len(pData):
pIntent = pData['Intent'][int(Idx[i])]
_, pModels = loadmodel(pRootDir, pAccountName, pIntent)
pPipeModel = make_pipeline(pVec, pModels)
pTe = TextExplainer(random_state=42).fit(
pData[pDesc][int(Idx[i])], pPipeModel.predict_proba)
pExplanation = pTe.explain_prediction()
pHtml = format_as_html(pExplanation,
force_weights=False,
include_styles=False,
horizontal_layout=True,
show_feature_values=False)
savehtml(pRootDir, pHtml, Idx[i], pIntent)
else:
print("Please select valid Id")
except Exception as e:
print(
'*** ERROR[003]: Error in visualiation file of eil5visual function: ',
sys.exc_info()[0], str(e))
print(traceback.format_exc())
return (-1)
return (0)
def main():
df = pd.read_excel('data/mr_vs_fr_30.xlsx')
df = df.sample(frac=1, random_state=seed)
df['text_lemmatized'] = df['text'].apply(morphText)
X_train, X_test, y_train, y_test = train_test_split(
df['text_lemmatized'], df['label'], test_size=0.3, random_state=42, stratify=df['label'])
flag_test = True
get_pipe(X_train, y_train, flag_test, X_test, y_test)
flag_test = False
pipe = get_pipe(df['text_lemmatized'], df['label'], flag_test)
k = 0
words = []
for index, row in df.iterrows():
te5 = TextExplainer(clf=DecisionTreeClassifier(max_depth=5), random_state=seed)
te5.fit(row['text_lemmatized'], pipe.predict_proba)
df_eli5_w = eli5.format_as_dataframe(te5.explain_weights())
print('class {}'.format('male' if row['label'] == 0 else 'woman'))
print('predict:')
print(df_eli5_w)
print(100*'*')
temp_m = ', '.join(df_eli5_w[df_eli5_w['weight'] > 0]['feature'].tolist())
if temp_m:
words.append(temp_m)
else:
words.append('')
k += 1
df['words'] = words
df.to_excel('mr_vs_fr_words_30.xlsx', index=False)
def test_lime_explain_probabilistic(newsgroups_train):
docs, y, target_names = newsgroups_train
try:
vec = HashingVectorizer(alternate_sign=False)
except TypeError:
# sklearn < 0.19
vec = HashingVectorizer(non_negative=True)
clf = MultinomialNB()
X = vec.fit_transform(docs)
clf.fit(X, y)
print(clf.score(X, y))
pipe = make_pipeline(vec, clf)
doc = docs[0]
te = TextExplainer(random_state=42)
te.fit(doc, pipe.predict_proba)
print(te.metrics_)
assert te.metrics_['score'] > 0.7
assert te.metrics_['mean_KL_divergence'] < 0.1
res = te.explain_prediction(top=20, target_names=target_names)
expl = format_as_text(res)
print(expl)
assert 'file' in expl
def _lime_analyze(self,
query,
indicies,
max_len,
max_replace,
top_targets=None):
model = self.model
vocab = self.vocab.word_to_idx
label = self.label.word_to_idx
prepro_query = self.preprocess(query)
explainer_generator = ExplainerGenerator(model, vocab, max_len)
sampler = MaskingTextSampler(replacement=UNK,
max_replace=max_replace,
token_pattern=None,
bow=False)
explainer_list = list()
for i in indicies:
predict_fn = explainer_generator.get_predict_function(i)
te = TextExplainer(
sampler=sampler,
position_dependent=True,
random_state=RANDOM_SEED,
)
te.fit(' '.join(prepro_query), predict_fn)
pred_explain = te.explain_prediction(
target_names=[l for l in label][3:], top_targets=top_targets)
explainer_list.append(pred_explain)
return explainer_list
def test_text_explainer_rbf_sigma():
text = 'foo bar baz egg spam'
predict_proba = substring_presence_predict_proba('bar')
te1 = TextExplainer().fit(text, predict_proba)
te2 = TextExplainer(rbf_sigma=0.1).fit(text, predict_proba)
te3 = TextExplainer(rbf_sigma=1.0).fit(text, predict_proba)
assert te1.similarity_.sum() < te3.similarity_.sum()
assert te1.similarity_.sum() > te2.similarity_.sum()
def test_text_explainer_token_pattern():
text = "foo-bar baz egg-spam"
predict_proba = substring_presence_predict_proba('bar')
# a different token_pattern
te = TextExplainer(token_pattern=r'(?u)\b[-\w]+\b')
te.fit(text, predict_proba)
print(te.metrics_)
assert te.metrics_['score'] > 0.95
assert te.metrics_['mean_KL_divergence'] < 0.1
expl = te.explain_prediction()
format_as_all(expl, te.clf_)
assert expl.targets[0].feature_weights.pos[0].feature == 'foo-bar'
def predict(model_id):
if os.path.exists("model/" + str(int(model_id)) + ".pkl"):
try:
if str(model_id) in clfs:
clf = clfs[str(model_id)]
else:
clf = joblib.load(model_file_name(model_id))
explainers = []
if is_text_type(model_id):
pipe = make_pipeline(vectorizer, clf)
prediction = pipe.predict(request.json)
for post in request.json:
te = TextExplainer(random_state=42, n_samples=500)
te.fit(post['text'], pipe.predict_proba)
made = te.explain_prediction(target_names=['pos', 'neg'])
explanation = made.targets[0].feature_weights
op_exp = {'pos': [], 'neg': []}
for feature in explanation.pos:
op_exp['pos'].append([feature.feature, feature.weight])
for feature in explanation.neg:
op_exp['neg'].append([feature.feature, feature.weight])
explainers.append(op_exp)
else:
rows = request.json
query = pd.get_dummies(pd.DataFrame(rows))
query = query.reindex(columns=model_columns, fill_value=0)
prediction = clf.predict(query)
for index, row in query.iterrows():
explanation = eli5.explain_prediction(
clf, row).targets[0].feature_weights
op_exp = {'pos': [], 'neg': []}
for feature in explanation.pos:
op_exp['pos'].append([feature.feature, feature.weight])
for feature in explanation.neg:
op_exp['neg'].append([feature.feature, feature.weight])
explainers.append(op_exp)
# Converting to int from int64
return jsonify({
"predictions": list(map(str, prediction)),
"explanations": explainers
})
except Exception as e:
return jsonify({'error': str(e), 'trace': traceback.format_exc()})
else:
print('train first')
return 'no model here'
def highlight_text(text):
predict_dict = predict(text)
try:
te = TextExplainer(random_state=42, n_samples=1000)
te.fit(text, nn_model.predict_proba)
highlight_html = te.show_prediction(
target_names=[val for val in CLASSES.values()],
top_targets=3,
top=200)
predict_dict["highlight"] = highlight_html
except:
predict_dict["highlight"] = None
return predict_dict
def explain(self,
doc,
truncate_len=512,
all_targets=False,
n_samples=2500):
"""
Highlights text to explain prediction
Args:
doc (str): text of documnet
truncate_len(int): truncate document to this many words
all_targets(bool): If True, show visualization for
each target.
n_samples(int): number of samples to generate and train on.
Larger values give better results, but will take more time.
Lower this value if explain is taking too long.
"""
is_array, is_pair = detect_text_format(doc)
if is_pair:
warnings.warn(
'currently_unsupported: explain does not currently support sentence pair classification'
)
return
if not self.c:
warnings.warn(
'currently_unsupported: explain does not support text regression'
)
return
try:
import eli5
from eli5.lime import TextExplainer
except:
msg = 'ktrain requires a forked version of eli5 to support tf.keras. '+\
'Install with: pip3 install git+https://github.com/amaiya/eli5@tfkeras_0_10_1'
warnings.warn(msg)
return
prediction = [self.predict(doc)] if not all_targets else None
if not isinstance(doc, str): raise Exception('text must of type str')
if self.preproc.is_nospace_lang():
doc = self.preproc.process_chinese([doc])
doc = doc[0]
doc = ' '.join(doc.split()[:truncate_len])
te = TextExplainer(random_state=42, n_samples=n_samples)
_ = te.fit(doc, self.predict_proba)
return te.show_prediction(target_names=self.preproc.get_classes(),
targets=prediction)
def limeTextClassification(
dataset, data, pr=Predictor(callingFunction="TextClassifier")
): # example retrieved from https://eli5.readthedocs.io/en/latest/tutorials/black-box-text-classifiers.html#textexplainer
pr.dataset = dataset
resultColumnName = pr.resultColumn
dataClasses = list(dict.fromkeys(data[resultColumnName].astype(str)))
dataClasses.sort()
te = TextExplainer(random_state=42)
def test_lime_flat_neighbourhood(newsgroups_train):
docs, y, target_names = newsgroups_train
doc = docs[0]
@_apply_to_list
def predict_proba(doc):
""" This function predicts non-zero probabilities only for 3 labels """
proba_graphics = [0, 1.0, 0, 0]
proba_other = [0.9, 0, 0.1, 0]
return proba_graphics if 'file' in doc else proba_other
te = TextExplainer(expand_factor=None, random_state=42)
te.fit(doc, predict_proba)
print(te.metrics_)
print(te.clf_.classes_, target_names)
res = te.explain_prediction(top=20, target_names=target_names)
for expl in format_as_all(res, te.clf_):
assert 'file' in expl
assert "comp.graphics" in expl
def st_lime_explanation(
text: str,
predict_func: Callable[[List[str]], np.ndarray],
unique_labels: List[str],
n_samples: int,
position_dependent: bool = True,
):
# TODO just use ELI5's built-in visualization when streamlit supports it:
# https://github.com/streamlit/streamlit/issues/779
with st.spinner("Generating LIME explanations..."):
te = TextExplainer(
random_state=1, n_samples=n_samples, position_dependent=position_dependent
)
te.fit(text, predict_func)
st.json(te.metrics_)
explanation = te.explain_prediction()
explanation_df = eli5.format_as_dataframe(explanation)
for target_ndx, target in enumerate(
sorted(explanation.targets, key=lambda t: -t.proba)
):
target_explanation_df = explanation_df[
explanation_df["target"] == target_ndx
].copy()
target_explanation_df["contribution"] = (
target_explanation_df["weight"] * target_explanation_df["value"]
)
target_explanation_df["abs_contribution"] = abs(
target_explanation_df["contribution"]
)
target_explanation_df = (
target_explanation_df.drop("target", axis=1)
.sort_values(by="abs_contribution", ascending=False)
.reset_index(drop=True)
)
st.subheader(
f"Target: {unique_labels[target_ndx]} (probability {target.proba:.4f}, score {target.score:.4f})"
)
st.dataframe(target_explanation_df)
def limeTextClassification(
dataset, data, pr=Predictor(callingFunction="TextClassifier")
): # example retrieved from https://eli5.readthedocs.io/en/latest/tutorials/black-box-text-classifiers.html#textexplainer
pr = Predictor(dataset=dataset, callingFunction="TextClassifier")
resultColumnName = pr.resultColumn
dataClasses = list(dict.fromkeys(data[resultColumnName].astype(str)))
dataClasses.sort()
pr = Predictor(dataset=dataset, callingFunction="TextClassifier")
te = TextExplainer(random_state=42)
te.fit(dataset["text"], pr.predict_proba)
te.fit(dataset["text"], pr.predict_proba)
te.show_prediction(target_names=pr._classes_000.tolist())
return te, pr._classes_000.tolist()
'''
def test_text_explainer_custom_classifier():
text = "foo-bar baz egg-spam"
predict_proba = substring_presence_predict_proba('bar')
# use decision tree to explain the prediction
te = TextExplainer(clf=DecisionTreeClassifier(max_depth=2))
te.fit(text, predict_proba)
print(te.metrics_)
assert te.metrics_['score'] > 0.99
assert te.metrics_['mean_KL_divergence'] < 0.01
expl = te.explain_prediction()
format_as_all(expl, te.clf_)
# with explain_weights we can get a nice tree representation
expl = te.explain_weights()
print(expl.decision_tree.tree)
assert expl.decision_tree.tree.feature_name == "bar"
format_as_all(expl, te.clf_)
def test_text_explainer_char_based(token_pattern):
text = "Hello, world!"
predict_proba = substring_presence_predict_proba('lo')
te = TextExplainer(char_based=True, token_pattern=token_pattern)
te.fit(text, predict_proba)
print(te.metrics_)
assert te.metrics_['score'] > 0.95
assert te.metrics_['mean_KL_divergence'] < 0.1
res = te.explain_prediction()
format_as_all(res, te.clf_)
check_targets_scores(res)
assert res.targets[0].feature_weights.pos[0].feature == 'lo'
# another way to look at results (not that useful for char ngrams)
res = te.explain_weights()
assert res.targets[0].feature_weights.pos[0].feature == 'lo'
def test_text_explainer_show_methods():
pytest.importorskip('IPython')
from IPython.display import HTML
text = "Hello, world!"
@_apply_to_list
def predict_proba(doc):
return [0.0, 1.0] if 'lo' in doc else [1.0, 0.0]
te = TextExplainer()
te.fit(text, predict_proba)
pred_expl = te.show_prediction()
assert isinstance(pred_expl, HTML)
assert 'lo' in pred_expl.data
weight_expl = te.show_weights()
assert isinstance(weight_expl, HTML)
assert 'lo' in weight_expl.data
train = fetch_subset('train')
test = fetch_subset('test')
vec = TfidfVectorizer(min_df=3, stop_words='english', ngram_range=(1, 2))
svd = TruncatedSVD(n_components=100, n_iter=7, random_state=42)
lsa = make_pipeline(vec, svd)
clf = SVC(C=150, gamma=2e-2, probability=True)
pipe = make_pipeline(lsa, clf)
pipe.fit(twenty_train.data, twenty_train.target)
pipe.score(twenty_test.data, twenty_test.target)
doc = twenty_test.data[0]
print_prediction(doc)
te = TextExplainer(random_state=42)
te.fit(doc, pipe.predict_proba)
#print(te.explain_prediction(target_names=twenty_train.target_names))
#print(eli5.format_as_image(te.explain_weights(target_names=twenty_train.target_names)))
show_html = lambda html: display(HTML(html))
show_html_expl = lambda expl, **kwargs: show_html(
format_as_html(expl, include_styles=False, **kwargs))
show_html(format_html_styles())
weights = eli5.show_weights(clf,
vec=vec,
target_names=train['target_names'],
horizontal_layout=False)
pred = show_html_expl(explain_prediction(clf,
__author__ = 'xead'
# coding: utf-8
from sentiment_classifier import SentimentClassifier
from sklearn.externals import joblib
from eli5.lime import TextExplainer
#clf = SentimentClassifier()
#pred = clf.get_prediction_message("Хороший телефон")
text = 'Хороший был у меня телефон 5 лет назад'
pipe = joblib.load("./pipe6.pkl")
te = TextExplainer(random_state=42)
te.fit(text, pipe.predict_proba)
res = te.show_prediction(target_names=['negative', 'positive'], top=25)
print (res)
# + Data
# + Model
# + Target Names
# + Function
# In[499]:
from eli5.lime import TextExplainer
# In[500]:
pipe.predict_proba
# In[501]:
exp = TextExplainer(random_state=42)
# In[502]:
X_test.values[0]
# In[515]:
a = pipelog.predict([input()])
if a == 1:
print("hate statement")
elif a == 0:
print("Not hate bro!")
# In[374]:
# -*- coding: utf-8 -*-
"""
__title__ = 'eli5'
__author__ = 'JieYuan'
__mtime__ = '2018/8/21'
"""
from eli5.lime import TextExplainer
from sklearn.feature_extraction.text import TfidfVectorizer
from sklearn.linear_model import LogisticRegression
from sklearn.pipeline import make_pipeline
X = [
"The dimension of the input documents is reduced to 100, and then a kernel SVM is used to classify the documents.",
"This is what the pipeline returns for a document - it is pretty sure the first message in test data belongs to sci.med:"
]
y = [0, 1]
piplie = make_pipeline(TfidfVectorizer(), LogisticRegression())
te = TextExplainer(random_state=42)
te.fit(X[0], piplie.predict_proba)
te.show_prediction()
te.show_weights()
eli5.show_prediction
for word in words:
index = 0
for word_block in word:
if len(average_word_vector) == index:
average_word_vector.append(0)
average_word_vector[index] += float(word_block)
index += 1
index = 0
for word_block in average_word_vector:
average_word_vector[index] /= float(len(words))
index += 1
xout.append(average_word_vector)
return np.array(xout)
vectorizer = V()
for classifier in classifiers:
print(classifier)
gnb = classifier()
pipe = make_pipeline(vectorizer, gnb)
pipe.fit(x[:testcutoff], y[:testcutoff])
y_predicted = pipe.predict_proba(x[testcutoff:])
#print(classification_report(y[testcutoff:], y_predicted, target_names=['known weird', 'less weird']))
te = TextExplainer(random_state=101, n_samples=500)
te.fit('Green new deal is the best bro, bring it on', pipe.predict_proba)
te.show_prediction(target_names=['known weird', 'less weird'])
x_train, x_test, y_train, y_test = train_test_split(X,Y, test_size = 0.1, random_state = 40, stratify=Y)
text_model.fit(x_train, y_train)
text_model.score(x_test, y_test)
from IPython.display import display, HTML
import eli5
from eli5.lime import TextExplainer
for idx in x_test.index[190:210]:
te = TextExplainer(random_state=42)
te.fit(cleaner(x_test[idx]), text_model.predict_proba, )
print("Real Class:", ["Non Toxic" if x == 0 else "Toxic" for x in [df_corpus_final.iloc[idx]['class']]])
print("Text uncleaned tweet:", df_corpus_final.iloc[idx]['tweet'])
print("ELI5 Predicted Class:")
HTML(display((te.show_prediction(target_names=[ 'Non Toxic','Toxic',]))))
import pickle
pickle.dump(text_model, open('toxic.pickle', 'wb'))
def get_result(self, text):
te = TextExplainer(random_state=42)
te.fit(text, self.pipe.predict_proba)
res = te.show_prediction(target_names=['negative', 'positive'], top=25)
return res
# Used in pickle pipeline on TF-IDF
def dummy(token):
return token
# Load pre-trained ML model
model = pickle.load(open('model.pkl', 'rb')) # NEEDS TO BE CREATED WITH BOTH FILES IN FOLDER pickle_model_for_webapp
# Create object of class preprocessing to clean data
reading = clean_data.preprocessing.preprocessing(convert_lower=True, use_spell_corrector=True, only_verbs_nouns=False)
# clf: define ML classifier
# vec: define vectorizer
# n_samples: sets the number of random examples to generate from given instance of text (default value 5000)
# use LIME method to train a white box classifier to make the same prediction as the black box one (pipeline)
te = TextExplainer(vec=TfidfVectorizer(ngram_range=(1, 2), preprocessor=dummy, token_pattern='(?u)\\b\\w+\\b'),
n_samples=5000, char_based=False, random_state=42)
def one_word_get_prediction_class_name(prediction):
'''
Pipeline with XGBoost - translate the prediction class number into words
:param prediction: the predicted number/class
:return: the predicted class in natural language
'''
# The order of classes in predict_proba: ['hate speech', 'neither', 'offensive language']
if prediction == 0:
output = "hate speech"
elif prediction == 1:
output = "neither"
else:
output = "offensive language"
def test_text_explainer_position_dependent():
text = "foo bar baz egg spam bar baz egg spam ham"
@_apply_to_list
def predict_proba(doc):
tokens = doc.split()
# 'bar' is only important in the beginning of the document,
# not in the end
return [0, 1] if len(tokens) >= 2 and tokens[1] == 'bar' else [1, 0]
# bag of words model is not powerful enough to explain predict_proba above
te = TextExplainer(random_state=42, vec=CountVectorizer())
te.fit(text, predict_proba)
print(te.metrics_)
assert te.metrics_['score'] < 0.9
assert te.metrics_['mean_KL_divergence'] > 0.3
# position_dependent=True can make it work
te = TextExplainer(position_dependent=True, random_state=42)
te.fit(text, predict_proba)
print(te.metrics_)
assert te.metrics_['score'] > 0.95
assert te.metrics_['mean_KL_divergence'] < 0.3
expl = te.explain_prediction()
format_as_all(expl, te.clf_)
# it is also possible to almost make it work using a custom vectorizer
vec = CountVectorizer(ngram_range=(1, 2))
te = TextExplainer(vec=vec, random_state=42)
te.fit(text, predict_proba)
print(te.metrics_)
assert te.metrics_['score'] > 0.95
assert te.metrics_['mean_KL_divergence'] < 0.3
expl = te.explain_prediction()
format_as_all(expl, te.clf_)
# custom vectorizers are not supported when position_dependent is True
with pytest.raises(ValueError):
te = TextExplainer(position_dependent=True, vec=HashingVectorizer())
a_set = Sentence(p_str)
stacked_embeddings.embed(a_set)
to_ret = a_set.get_embedding().cpu().detach().numpy(
).reshape(1, -1)
except:
print(type(X))
print(X)
return to_ret
pipe = joblib.load('saved_card_classification.pkl')
if keras:
pipe.named_steps['model'].model = load_model('keras_model.h5')
te = TextExplainer(random_state=42, n_samples=3000, position_dependent=False)
def explain_pred(sentence):
te.fit(sentence, pipe.predict_proba)
#txt = format_as_text(te.explain_prediction(target_names=["green", "neutral", "red"]))
t_pred = te.explain_prediction(top=20,
target_names=[
"ANB",
"CAP",
"ECON",
"EDU",
"ENV",
"EX",
"FED",
"HEG",
import jieba
seg_list = jieba.cut("看了快一半了才发现是mini的广告", cut_all=False)
list(seg_list)
# ### Example 1
# In[10]:
get_proba(["看 了 快 一半 了 才 发现 是 mini 的 广告"])
# In[11]:
from eli5.lime import TextExplainer
te = TextExplainer(random_state=42, n_samples=5000)
te.fit(" ".join(jieba.cut("看了快一半了才发现是mini的广告", cut_all=False)), get_proba)
te.show_prediction(target_names=["neg", "pos"])
# In[12]:
te.metrics_
# In[13]:
te.samples_[:10]
# #### Character-based Whitebox
# In[14]:
# opcodes_dir = '/home/hwangdz/coreutils/coreutils-8.28/install_m32/bin/md5funcs_ops'
opcodes_dir = '/home/hwangdz/git/rl-select-div/only-similarity/explanation/%s_ops_info' % bin_name
output_dir = 'explanation/%s_html' % bin_name
if not os.path.isdir(output_dir):
os.mkdir(output_dir)
for file_name in os.listdir(opcodes_dir):
# if file_name != 'dump.s':
# continue
if file_name == 'op_distribution':
continue
file_path = os.path.join(opcodes_dir, file_name)
with open(file_path, 'r') as f:
op_codes = f.read()
if len(op_codes) < 20:
continue
num_ops = len(op_codes.split())
op_codes = op_codes.replace('\n', ' ')
opcode_explainer = TextExplainer(random_state=59, sampler=ops_sampler, n_samples=5000)
#repeat_times = (len(op_codes.split()) / 100) ** 2
repeat_times = 1
for _ in range(repeat_times):
opcode_explainer.fit(op_codes, ss.predict_proba)
explanation = opcode_explainer.explain_prediction()._repr_html_()
with open('explanation/%s_html/explanation-%s.html' % (bin_name, file_name), 'w') as ef:
ef.write(explanation)
ef.write('num of opcodes: %d\n' % num_ops)
ef.write('</br>\n')
ef.write(op_codes)
