def test_binary(self):
"""
:return:
"""
cols = ['C1', 'D', 'E', 'F']
X = self.create_dataset(n_rows=1000)
X_t = self.create_dataset(n_rows=100)
enc = encoders.BinaryEncoder(verbose=1, cols=cols)
enc.fit(X, None)
self.verify_numeric(enc.transform(X_t))
enc = encoders.BinaryEncoder(verbose=1)
enc.fit(X, None)
self.verify_numeric(enc.transform(X_t))
enc = encoders.BinaryEncoder(verbose=1, drop_invariant=True)
enc.fit(X, None)
self.verify_numeric(enc.transform(X_t))
enc = encoders.BinaryEncoder(verbose=1, return_df=False)
enc.fit(X, None)
self.assertTrue(isinstance(enc.transform(X_t), np.ndarray))
def encoder(data):
###### BINARY ENCODER ######'
B_encoder = category_encoders.BinaryEncoder(cols=['premis_var'])
data = B_encoder.fit_transform(data)
B_encoder = category_encoders.BinaryEncoder(cols=['incdt_time'])
data = B_encoder.fit_transform(data)
B_encoder = category_encoders.BinaryEncoder(cols=['incdt_date'])
data = B_encoder.fit_transform(data)
###### DATA STANDARDISATION ######
# sc = StandardScaler()
# data['incdt_time'] = sc.fit_transform(data['incdt_time'].values.reshape(-1,1))
# data['incdt_date'] = sc.fit_transform(data['incdt_date'].values.reshape(-1,1))
###### ONE HOT ENCODER ######
OH_encoder = OneHotEncoder()
hc1 = DataFrame(OH_encoder.fit_transform(data['BORO_NM'].values.reshape(-1,1)).toarray(),
columns = ['BORO1', 'BORO2','BORO3', 'BORO4', 'BORO5'])
hc2 = DataFrame(OH_encoder.fit_transform(data['VIC_AGE_GROUP'].values.reshape(-1, 1)).toarray(),
columns=['VIC_AGE1', 'VIC_AGE2', 'VIC_AGE3', 'VIC_AGE4', 'VIC_AGE5'])
hc3 = DataFrame(OH_encoder.fit_transform(data['LOC_OF_OCCUR_DESC'].values.reshape(-1, 1)).toarray(),
columns=['LOC_DESC1', 'LOC_DESC2', 'LOC_DESC3', 'LOC_DESC4'])
hc4 = DataFrame(OH_encoder.fit_transform(data['VIC_RACE'].values.reshape(-1, 1)).toarray(),
columns=['VIC_RACE1', 'VIC_RACE2', 'VIC_RACE3', 'VIC_RACE4'])
hc5 = DataFrame(OH_encoder.fit_transform(data['VIC_SEX'].values.reshape(-1, 1)).toarray(),
columns=['VICM_SEX1', 'VICM_SEX2'])
hc6 = DataFrame(OH_encoder.fit_transform(data['SUSP_RACE'].values.reshape(-1, 1)).toarray(),
columns=['SUSP_RACE1', 'SUSP_RACE2', 'SUSP_RACE3', 'SUSP_RACE4'])
hc7 = DataFrame(OH_encoder.fit_transform(data['VIC_SEX'].values.reshape(-1, 1)).toarray(),
columns=['SUSP_SEX1', 'SUSP_SEX2'])
data = pd.concat([data,hc1,hc2,hc3,hc4,hc5,hc6,hc7], axis=1)
return data
def binary_encoding(df, cols, handle_nan=True):
if handle_nan:
encoder = ce.BinaryEncoder(cols=cols,
handle_unknown='indicator',
handle_missing='indicator')
else:
encoder = ce.BinaryEncoder(cols=cols,
handle_unknown='return_nan',
handle_missing='return_nan')
df_new = encoder.fit_transform(df)
return df_new
def process_data(data_df: pd.DataFrame) -> pd.DataFrame:
encoder = ce.BinaryEncoder(cols=['b', 'm', 'd', 'p', 'l', 's'],
return_df=True)
data_encoded_df = encoder.fit_transform(data_df)
print(data_encoded_df)
return data_encoded_df
def BinaryEncodingCalc(request, fName):
df = get_df(fName)
if request.method == 'POST':
selected_cols = request.POST.getlist('binaryCol')
for selected_col in selected_cols:
encoder = ce.BinaryEncoder(cols=[selected_col])
df = encoder.fit_transform(df)
df.to_csv(os.path.join(settings.MEDIA_ROOT,
'processed/' + fName + '.csv'),
index=False)
df_new = get_df(fName)
clm_list = list(df_new)
NaN_percent = get_NaN_percent(fName)
binary_list = []
for clm in clm_list:
dt = df_new[clm].dtype
if dt == 'int64' or dt == 'float64':
pass
else:
binary_list.append(clm)
binaryProcessed_list = selected_cols
context = {
'fName': fName,
'processing_list': binary_list,
'processed_list': binaryProcessed_list,
'NaN_percent': NaN_percent,
'status': 'Success',
'message': 'Binary Encoding was done on selected features.'
}
return render(request, 'BinaryEncoding.html', context)
def create_binary(df, col):
assert isinstance(col, str)
uniques = pd.DataFrame(df[col].unique(), columns=[col])
enc = ce.BinaryEncoder(verbose=1, cols=[col])
uniques = pd.concat([uniques, enc.fit_transform(uniques)], axis=1)
return df.merge(uniques, how='left', on=col)
def apply_binary_encoding(df, categorical_columns):
if not isinstance(df, pd.DataFrame):
raise DataFrameTypeError('df', df)
import category_encoders as ce
encoder = ce.BinaryEncoder(cols=categorical_columns).fit(df.values)
X_transformed = encoder.transform(df)
return X_transformed
def encode_categorical(given_dataset):
"""Street,CentralAir -> binary feature encoding"""
"""LotShape,LandContour,LandSlope,BldgType,HouseStyle,ExterQual,ExterCond,Foundation,BsmtQual
,BsmtCond,BsmtExposure,BsmtFinType1,BsmtFinType2,HeatingQC,Electrical
,KitchenQual,Functional,FireplaceQu,GarageType,GarageFinish,GarageQual
,GarageCond,PoolQC,Fence-> ordinal feature encoding """
"""MSZoning,Alley,Utilities,LotConfig,Neighborhood,Condition1,Condition2,RoofStyle,RoofMatl,Exterior1st
,Exterior2nd,MasVnrType,Heating,Functional,PavedDrive,MiscFeature,SaleType
,SaleCondition -> nominal feature"""
encoder1 = ce.BinaryEncoder(cols=['Street','CentralAir'], return_df=True)
encoder2 = ce.OrdinalEncoder(cols=['MSZoning','Alley','LandSlope','BldgType','HouseStyle','ExterQual'
,'ExterCond','Foundation','BsmtQual','BsmtCond'
,'BsmtExposure','BsmtFinType1','BsmtFinType2'
,'HeatingQC','Electrical','KitchenQual'
,'FireplaceQu','GarageType','GarageFinish'
,'GarageQual','GarageCond','PoolQC','Fence'], return_df=True)
encoder3 = ce.OneHotEncoder(cols=['LotShape','LandContour','Utilities','LotConfig','Neighborhood'
,'Condition1','Condition2','RoofStyle','RoofMatl','Exterior1st'
,'Exterior2nd','MasVnrType','Heating','PavedDrive'
,'MiscFeature','SaleType','SaleCondition','Functional'], return_df=True)
# Assume our loan data has been imported as df already
given_dataset = encoder1.fit_transform(given_dataset)
given_dataset = encoder2.fit_transform(given_dataset)
given_dataset = encoder3.fit_transform(given_dataset)
return given_dataset
def general_imputer(df):
"""
general imputer + normalizer and binarizer
"""
nulls_per_column = df.isnull().sum()
print(nulls_per_column.loc[nulls_per_column > 0])
# Create a boolean mask for categorical columns
categorical_feature_mask = df.dtypes == object
# Get list of categorical column names
feat_cat = df.columns[categorical_feature_mask].tolist()
# Get list of non-categorical column names
feat_cont = df.columns[~categorical_feature_mask].tolist()
mapper = DataFrameMapper(
[([nf],
Pipeline(steps=[('imp', SimpleImputer(
strategy="median")), ('std', StandardScaler())]))
for nf in feat_cont] + [(cf,
Pipeline(steps=[
('imp', CategoricalImputer()),
('label', ce.BinaryEncoder()),
])) for cf in feat_cat],
input_df=True,
df_out=True,
default=False)
return (mapper)
def test_inverse_transform_11(self):
"""
Test binary encoding
"""
train = pd.DataFrame({
'city': ['chicago', 'paris'],
'state': ['US', 'FR'],
'other': ['A', np.nan]
})
test = pd.DataFrame({
'city': ['chicago', 'paris', 'monaco'],
'state': ['US', 'FR', 'FR'],
'other': ['A', np.nan, 'B']
})
expected = pd.DataFrame({
'city': ['chicago', 'paris', np.nan],
'state': ['US', 'FR', 'FR'],
'other': ['A', np.nan, 'B']
})
enc = ce.BinaryEncoder(cols=['city', 'state']).fit(train)
result = enc.transform(test)
original = inverse_transform(result, enc)
pd.testing.assert_frame_equal(original, expected)
def featureEncodingUsingBinaryEncoder(dataSetForFeatureEncoding):
print("****** Start binary encoding on the categorical features in the given dataset *****")
labelName = getLabelName()
#Extract the categorical features, leave the label
categoricalColumnsInTheDataSet = dataSetForFeatureEncoding.drop([labelName],axis=1).select_dtypes(['object'])
#Get the names of the categorical features
categoricalColumnNames = categoricalColumnsInTheDataSet.columns.values
print("****** Number of features before binary encoding: ",len(dataSetForFeatureEncoding.columns))
print("****** Number of categorical features in the dataset: ",len(categoricalColumnNames))
print("****** Categorical feature names in the dataset: ",categoricalColumnNames)
print('\n****** Here is the list of unique values present in each categorical feature in the dataset *****\n')
label = dataSetForFeatureEncoding.drop(dataSetForFeatureEncoding.loc[:, ~dataSetForFeatureEncoding.columns.isin([labelName])].columns, axis = 1)
for feature in categoricalColumnNames:
uniq = np.unique(dataSetForFeatureEncoding[feature])
print('\n{}: {} '.format(feature,len(uniq)))
printList(dataSetForFeatureEncoding[feature].unique(),'distinct values')
featureColumns = dataSetForFeatureEncoding.drop(dataSetForFeatureEncoding.loc[:, ~dataSetForFeatureEncoding.columns.isin([feature])].columns, axis = 1)
binaryEncoder = ce.BinaryEncoder(cols = [feature])
binaryEncodedFeature = binaryEncoder.fit_transform(featureColumns, label)
dataSetForFeatureEncoding = dataSetForFeatureEncoding.join(binaryEncodedFeature)
dataSetForFeatureEncoding = dataSetForFeatureEncoding.drop(feature, axis=1)
dataSetForFeatureEncoding = dataSetForFeatureEncoding.drop(labelName, axis=1)
dataSetForFeatureEncoding[labelName] = label
print("****** Number of features after binary encoding: ",len(dataSetForFeatureEncoding.columns))
print("****** End binary encoding on the categorical features in the given dataset *****\n")
return dataSetForFeatureEncoding
def processingPreModelling(df, catVarsDict={}):
# Data Processing before Modelling
# Bools to int
df.isMobile = df.isMobile.astype(int)
# Date objects to datetime format
df.date = pd.to_datetime(df.date, format='%Y-%m-%d')
for var, value in catVarsDict.items():
if (value == 'BinaryEncoder'):
encoder = category_encoders.BinaryEncoder(cols=[var],
drop_invariant=True,
return_df=True)
df = encoder.fit_transform(df)
elif (value == 'LabelEncoder'):
df[var] = LabelEncoder().fit_transform(df[var])
elif (value == 'OneHot'):
encoder = category_encoders.one_hot.OneHotEncoder(
cols=[var],
drop_invariant=True,
return_df=True,
use_cat_names=True)
df = encoder.fit_transform(df)
return df
def binary_encoding():
source_bin_df = data_df
encoder = ce.BinaryEncoder(cols=['decision'], drop_invariant=True)
dfb = encoder.fit_transform(source_bin_df['decision'])
source_bin_df = pd.concat([source_bin_df, dfb], axis=1)
print(source_bin_df)
def encode_vars_via_lookup(fset, feature_lookup):
# import category_encoders as ce
for var in fset:
encoder = None
vtype = feature_lookup.get(var, 'numeric') # default numeric
if vtype == 'ord':
encoder = ce.OrdinalEncoder(cols=[
var,
])
elif vtype == 'cat':
encoder = ce.OneHotEncoder(cols=[
var,
])
elif vtype in ('str', 'high_card'
): # high_card: categorical but with high cardinality
encoder = ce.BinaryEncoder(cols=[
var,
]) # ... or use ce.HashingEncoder()
else:
# assuming that the var is numeric
pass
# data imputation [todo]
if encoder is not None:
dfX = encoder.fit_transform(dfX, dfy)
return dfX
def test_inv_transform_ct_9(self):
"""
test inv_transform_ct with Binary Encoder and passthrough option
"""
y = pd.DataFrame(data=[0, 1], columns=['y'])
train = pd.DataFrame({'city': ['chicago', 'paris'],
'state': ['US', 'FR'],
'other': ['A', 'B']})
enc = ColumnTransformer(
transformers=[
('binary', ce.BinaryEncoder(), ['city', 'state'])
],
remainder='passthrough')
enc.fit(train, y)
test = pd.DataFrame({'city': ['chicago', 'chicago', 'paris'],
'state': ['US', 'FR', 'FR'],
'other': ['A', 'B', 'C']})
expected = pd.DataFrame({'binary_city': ['chicago', 'chicago', 'paris'],
'binary_state': ['US', 'FR', 'FR'],
'other': ['A', 'B', 'C']})
result = pd.DataFrame(enc.transform(test))
result.columns = ['col1_0', 'col1_1', 'col2_0', 'col2_1', 'other']
original = inverse_transform(result, enc)
pd.testing.assert_frame_equal(original, expected)
def dicho_nominales(data, cols=['var1', 'var2']):
'''
# Dichotomise les variables catégorielles nominales
'''
enc = ce.BinaryEncoder(cols=cols)
data_drop = enc.fit_transform(data)
return (data_drop)
def binary():
X, _, _ = get_mushroom_data()
print(X.info())
enc = ce.BinaryEncoder()
enc.fit(X, None)
out = enc.transform(X)
print(out.info())
del enc, _, X, out
def fit(self, X, y=None):
# we need the fit statement to accomodate the sklearn pipeline
des_nom_DF = X[self.descreteVars_Nominal]
# Map Nominal Categorical data to Numerical
cat_nom_DF = des_nom_DF.fillna('NULL').astype(str)
self.ce_binary = ce.BinaryEncoder()
self.ce_binary.fit(cat_nom_DF)
return self
def test_inverse_transform_ce_binary(self):
"""
Unit test inverse transform ce binary
"""
preprocessing = ce.BinaryEncoder(cols=['Age', 'Sex'], return_df=True)
fitted_dataset = preprocessing.fit_transform(self.ds_titanic_clean)
output = inverse_transform(fitted_dataset, preprocessing)
pd.testing.assert_frame_equal(output, self.ds_titanic_clean)
def encode_categorical_features(df, encoder=None):
'''
Encode the categorical features using BinaryEncoder.
'''
if encoder == None:
encoder = ce.BinaryEncoder(
cols=[location_key, class_key, subdepartment_key])
encoder.fit(df)
return encoder.transform(df), encoder
def create_features(self, df_train, df_test):
encoder = ce.BinaryEncoder(cols=self.columns)
encoder.fit(df_train[self.columns],
df_train[self.target_column].values.tolist())
encoded_train = encoder.transform(df_train[self.columns])
encoded_test = encoder.transform(df_test[self.columns])
for column in encoded_train.columns:
self.train[column + '_BinaryEncoder'] = encoded_train[column]
self.test[column + '_BinaryEncoder'] = encoded_test[column]
def load_and_clean(non_categorical,
categorical,
data_path="../data/with_stock_data_webclicks_linkedin.csv",
normalize=False,
binary_encode=False,
trend_features=True,
filter=False):
frame = pd.read_csv(data_path)
all_non_categorical = get_all_non_categorical()
all_categorical = get_all_categorical()
for non_c in non_categorical:
if non_c not in all_non_categorical:
raise ValueError("Non categorical {} is not valid".format(non_c))
for c in categorical:
if c not in all_categorical:
raise ValueError("Categorical {} is not valid".format(non_c))
X = pd.DataFrame({key: frame[key] for key in non_categorical})
if trend_features:
trends_data = google_trends_features()
X = pd.concat([X, google_trends_features()], axis=1)
y = frame[[Y]]
for cat in categorical:
new_cols = None
if binary_encode:
ce_binary = ce.BinaryEncoder(cols=[cat])
new_cols = ce_binary.fit_transform(frame[cat])
else:
new_cols = pd.get_dummies(frame[cat], prefix='category')
X = pd.concat([X, new_cols], axis=1)
X = clean_data(X)
remove_nan(X, non_categorical)
if filter:
indices = X["current employee estimate"] > 300
X = X[indices]
y = y[indices]
if (normalize):
X.loc[:, :] = preprocessing.StandardScaler().fit_transform(X)
X_train, X_test, y_train, y_test = train_test_split(X,
y,
test_size=0.3,
random_state=1)
X_dev, X_test, y_dev, y_test = train_test_split(X_test,
y_test,
test_size=.5,
random_state=1)
return (X_train, y_train, X_dev, y_dev, X_test, y_test)
def encode_categorical(df,encoder_name='binary'):
encoder_dic = {"one_hot":ce.OneHotEncoder(),
"feature_hashing":ce.HashingEncoder(n_components=32),
"binary":ce.BinaryEncoder(),
"ordinal":ce.OrdinalEncoder(),
"polynomial":ce.PolynomialEncoder()}
encoder = encoder_dic.get(encoder_name)
encoder.fit(df,verbose=1)
df = encoder.transform(df)
return df
def main(params, inputs, outputs):
columns_param = params.columns
data = inputs.data
data_new = outputs.data_new
data_0 = pd.read_pickle(data)
encoder = ce.BinaryEncoder(cols=[col for col in columns_param.split(",")])
data_1 = encoder.fit_transform(data_0)
data_1.to_pickle(data_new)
def test_binary_np(self):
"""
:return:
"""
X = self.create_array(n_rows=1000)
X_t = self.create_array(n_rows=100)
enc = encoders.BinaryEncoder(verbose=1)
enc.fit(X, None)
self.verify_numeric(enc.transform(X_t))
def categorical_cols_train(data):
try:
print('Categorical Encoding')
encoder = ce.BinaryEncoder(cols=[
'employment_type', 'required_experience', 'required_education',
'country'
])
newdata = encoder.fit_transform(data)
pickle.dump(encoder, open("model/encoder.p", "wb"))
return newdata
except Exception as e:
handle('categorical column handling')
def preprocess(self) -> DataFrame:
X_train = self.transaction_train_df_raw.drop('isFraud',
axis=1) # 506691
# num_test = 0.20
# X_all = transaction_train_df_raw.drop('isFraud', axis=1)
# Y_all = transaction_train_df_raw['isFraud']
# X_train, X_test, Y_train, Y_test = train_test_split(X_all, Y_all, test_size=num_test)
# Preprocessing for numerical data
numerical_transformer = SimpleImputer(strategy='mean')
# Preprocessing for categorical data
# to implment addr2, divide into 2 categories, then one hot enconding
onehot_categorical_transformer = Pipeline(
steps=[('imputer', SimpleImputer(
strategy='most_frequent')), ('onehot', ce.OneHotEncoder())])
binary_categorical_transformer = Pipeline(
steps=[('imputer', SimpleImputer(strategy='most_frequent')
), ('binary', ce.BinaryEncoder(drop_invariant=False))])
ordinary_categorical_transformer = Pipeline(
steps=[('imputer', SimpleImputer(
strategy='mean')), ('ordinary', ce.OrdinalEncoder())])
# Bundle preprocessing for numerical and categorical data
preprocessor_pipeline = ColumnTransformer(transformers=[
('num', numerical_transformer, [
'TransactionDT', 'TransactionAmt', 'C1', 'C2', 'C3', 'C4',
'C5', 'C6', 'C7', 'C8', 'C9', 'C10', 'C11', 'C12', 'C13',
'C14', 'D1', 'D4', 'D10', 'D15'
]),
('onehot_cat', onehot_categorical_transformer,
['ProductCD', 'card4', 'M6']),
# ('onehot_cat', onehot_categorical_transformer, ['ProductCD', 'card4', 'card6', 'M6']),
# binary encoder did not work, should re implement
# ('binary_cat', binary_categorical_transformer, ['card3', 'card5', 'addr1', 'addr2']),
('binary_cat', binary_categorical_transformer, ['P_emaildomain']),
('ordinary_cat', ordinary_categorical_transformer,
['card1', 'card2'])
])
# dataFrameMapper = DataFrameMapper([
# (['TransactionDT', 'TransactionAmt', 'C1', 'C2', 'C3', 'C4', 'C5', 'C6', 'C7', 'C8', 'C9', 'C10', 'C11', 'C12', 'C13', 'C14', 'D1', 'D4', 'D10', 'D15'], numerical_transformer),
# (['ProductCD', 'card4', 'M6'], onehot_categorical_transformer),
# # ('onehot_cat', onehot_categorical_transformer, ['ProductCD', 'card4', 'card6', 'M6']),
# # binary encoder did not work, should re implement
# # ('binary_cat', binary_categorical_transformer, ['card3', 'card5', 'addr1', 'addr2']),
# (['P_emaildomain'], binary_categorical_transformer ),
# (['card1', 'card2'], ordinary_categorical_transformer)
# ]
# )
return preprocessor_pipeline.fit_transform(X_train)
def test_binary(self):
"""
:return:
"""
cols = ['C1', 'D', 'E', 'F']
X = self.create_dataset(n_rows=1000)
X_t = self.create_dataset(n_rows=100)
enc = encoders.BinaryEncoder(verbose=1, cols=cols)
enc.fit(X, None)
self.verify_numeric(enc.transform(X_t))
enc = encoders.BinaryEncoder(verbose=1)
enc.fit(X, None)
self.verify_numeric(enc.transform(X_t))
enc = encoders.BinaryEncoder(verbose=1, drop_invariant=True)
enc.fit(X, None)
self.verify_numeric(enc.transform(X_t))
enc = encoders.BinaryEncoder(verbose=1, return_df=False)
enc.fit(X, None)
self.assertTrue(isinstance(enc.transform(X_t), np.ndarray))
# test inverse_transform
X = self.create_dataset(n_rows=1000, has_none=False)
X_t = self.create_dataset(n_rows=100, has_none=False)
X_t_extra = self.create_dataset(n_rows=100,
extras=True,
has_none=False)
enc = encoders.BinaryEncoder(verbose=1)
enc.fit(X, None)
self.verify_numeric(enc.transform(X_t))
self.verify_inverse_transform(
X_t, enc.inverse_transform(enc.transform(X_t)))
with self.assertRaises(ValueError):
out = enc.inverse_transform(enc.transform(X_t_extra))
def __init__(self, encoder_type, columns_name=None):
"""
:param encoder_type:
:param columns_name: list， 特征名组成的列表名
"""
if encoder_type == "BackwardDe": # 反向差分编码
self.encoder = ce.BackwardDifferenceEncoder(cols=columns_name)
elif encoder_type == "BaseN": # BaseN编码
self.encoder = ce.BaseNEncoder(cols=columns_name)
elif encoder_type == "Binary": # 二值编码
self.encoder = ce.BinaryEncoder(cols=columns_name)
elif encoder_type == "Catboost":
self.encoder = ce.CatBoostEncoder(cols=columns_name)
elif encoder_type == "Hash":
self.encoder = ce.HashingEncoder(cols=columns_name)
elif encoder_type == "Helmert":
self.encoder = ce.HelmertEncoder(cols=columns_name)
elif encoder_type == "JamesStein":
self.encoder = ce.JamesSteinEncoder(cols=columns_name)
elif encoder_type == "LOO": # LeaveOneOutEncoder 编码
self.encoder = ce.LeaveOneOutEncoder(cols=columns_name)
elif encoder_type == "ME":
self.encoder = ce.MEstimateEncoder(cols=columns_name) # M估计编码器
elif encoder_type == "OneHot":
self.encoder = ce.OneHotEncoder(cols=columns_name)
elif encoder_type == "OridinalEncoder": # 原始编码
self.encoder = ce.OrdinalEncoder(cols=columns_name)
elif encoder_type == "Sum": # 求和编码
self.encoder = ce.SumEncoder(cols=columns_name)
elif encoder_type == "Polynomial": # 多项式编码
self.encoder = ce.PolynomialEncoder(cols=columns_name)
elif encoder_type == "Target": # 目标编码
self.encoder = ce.TargetEncoder(cols=columns_name)
elif encoder_type == "WOE": # WOE 编码器
self.encoder = ce.WOEEncoder(cols=columns_name)
else:
raise ValueError("请选择正确的编码方式")
def main():
cat_columns = [GENDER, HAIR_COLOR, WEARS_GLASSES, PROFESSION, COUNTRY]
training = read_data(
"tcd ml 2019-20 income prediction training (with labels).csv")
training = preprocess_data(training, True)
# print(training.head())
testing = read_data(
"tcd ml 2019-20 income prediction test (without labels).csv")
testing = preprocess_data(testing, False)
training = pd.get_dummies(training,
columns=[GENDER, HAIR_COLOR, WEARS_GLASSES])
ce_bin = ce.BinaryEncoder(cols=[PROFESSION, COUNTRY])
training = ce_bin.fit_transform(training)
cat_dummies = [
col for col in training
if "_" in col and col.split("_")[0] in cat_columns
]
# print(training.dtypes)
# print(cat_dummies)
processed_cols = list(training.columns[:])
# testing = pd.get_dummies(testing, columns=cat_columns)
testing = pd.get_dummies(testing,
columns=[GENDER, HAIR_COLOR, WEARS_GLASSES])
testing = ce_bin.fit_transform(testing)
for col in testing.columns:
if ("_" in col) and (col.split("_")[0]
in cat_columns) and col not in cat_dummies:
print("Removing additional feature {}".format(col))
testing.drop(col, axis=1, inplace=True)
for col in cat_dummies:
if col not in testing.columns:
print("Adding missing feature {}".format(col))
testing[col] = 0
X = training.loc[:, training.columns != INCOME]
X = X.loc[:, X.columns != "Instance"]
y = training[INCOME]
process_data(testing, training)
