huge progress with getting feature names out from One Hot encoder
This commit is contained in:
parent
95852fa40e
commit
b49c877f49
1 changed files with 54 additions and 48 deletions
|
|
@ -13,6 +13,8 @@ categorical_ix = X.select_dtypes(include=['object', 'bool']).columns
|
|||
categorical_ix
|
||||
|
||||
# Determine preprocessing steps ~ var_type
|
||||
var_type = 'mixed'
|
||||
|
||||
if var_type == 'numerical':
|
||||
t = [('num', MinMaxScaler(), numerical_ix)]
|
||||
|
||||
|
|
@ -23,47 +25,52 @@ if var_type == 'mixed':
|
|||
t = [('cat', OneHotEncoder(), categorical_ix)
|
||||
, ('num', MinMaxScaler(), numerical_ix)]
|
||||
|
||||
t = [('num', MinMaxScaler(), numerical_ix)
|
||||
, ('cat', OneHotEncoder(), categorical_ix)]
|
||||
|
||||
col_transform = ColumnTransformer(transformers = t
|
||||
, remainder='passthrough')
|
||||
#--------------ALEX help
|
||||
# col_transform
|
||||
# col_transform.fit(X)
|
||||
# test = col_transform.transform(X)
|
||||
# print(col_transform.get_feature_names_out())
|
||||
|
||||
# foo = col_transform.fit_transform(X)
|
||||
# (foo == test).all()
|
||||
#-----------------------
|
||||
|
||||
col_transform.fit(X)
|
||||
col_transform.get_feature_names_out()
|
||||
|
||||
var_type_colnames = col_transform.get_feature_names_out()
|
||||
var_type_colnames = pd.Index(var_type_colnames)
|
||||
|
||||
if var_type == 'mixed':
|
||||
print('\nVariable type is:', var_type
|
||||
, '\nNo. of columns in input_df:', len(input_df.columns)
|
||||
, '\nNo. of columns post one hot encoder:', len(var_type_colnames))
|
||||
else:
|
||||
print('\nNo. of columns in input_df:', len(input_df.columns))
|
||||
|
||||
|
||||
# %% begin stupid
|
||||
stupid=OneHotEncoder()
|
||||
stupid.fit(X[categorical_ix])
|
||||
stupid_thing = stupid.get_feature_names()
|
||||
horrid = (list(stupid_thing) + list(numerical_ix))
|
||||
# stupid = OneHotEncoder()
|
||||
# stupid.fit(X[categorical_ix])
|
||||
# stupid_thing = stupid.get_feature_names()
|
||||
# print(len(stupid_thing))
|
||||
# horrid = (list(stupid_thing) + list(numerical_ix))
|
||||
# print(horrid)
|
||||
|
||||
asdfasdf = pd.Index(horrid)
|
||||
# print(len(horrid))
|
||||
# asdfasdf = pd.Index(horrid)
|
||||
|
||||
asdfasdf[gscv_fs.best_estimator_.named_steps['fs'].get_support()]
|
||||
|
||||
|
||||
col_transform.get_param_names()['transformers']
|
||||
|
||||
len(stupid.get_feature_names())
|
||||
len(numerical_ix)
|
||||
|
||||
|
||||
# cat_trans = Pipeline(steps=[('onehot',OneHotEncoder(), categorical_ix)])
|
||||
# num_trans = Pipeline(steps=[('num', MinMaxScaler(), numerical_ix)])
|
||||
|
||||
# pre_p = ColumnTransformer(transformers = [('num', num_trans, numerical_ix),
|
||||
# ('cat', cat_trans, categorical_ix)
|
||||
# ]
|
||||
|
||||
# annoying = Pipeline([('preprocessor', pre_p),('clf', LogisticRegression())])
|
||||
|
||||
# fukkit = GridSearchCV(annoying
|
||||
# , search_space
|
||||
# , cv = cv
|
||||
# , scoring = mcc_score_fn
|
||||
# , refit = 'mcc'
|
||||
# , verbose = 1
|
||||
# , return_train_score = True
|
||||
# , **njobs)
|
||||
# fukkit.fit(X, y)
|
||||
# fukkit.best_params_
|
||||
# fukkit.best_score_
|
||||
# asdfasdf[gscv_fs.best_estimator_.named_steps['fs'].get_support()]
|
||||
|
||||
# col_transform.get_param_names()['transformers']
|
||||
|
||||
# len(stupid.get_feature_names())
|
||||
# len(numerical_ix)
|
||||
|
||||
# end stupid
|
||||
#%%
|
||||
|
|
@ -77,12 +84,11 @@ pipe = Pipeline([
|
|||
#cv = rskf_cv
|
||||
cv = skf_cv
|
||||
|
||||
# my data: Feature Selelction + GridSearch CV + Pipeline
|
||||
|
||||
# LR: Feature Selelction + GridSearch CV + Pipeline
|
||||
search_space = [
|
||||
{ 'fs__estimator': [LogisticRegression(**rs)]
|
||||
, 'fs__min_features_to_select': [0,1]
|
||||
,'fs__cv': [rskf_cv]
|
||||
, 'fs__min_features_to_select': [1]
|
||||
,'fs__cv': [skf_cv]
|
||||
},
|
||||
{
|
||||
#'clf': [LogisticRegression()],
|
||||
|
|
@ -108,7 +114,7 @@ search_space = [
|
|||
# 'clf__n_neighbors': [3, 7, 11],
|
||||
# 'clf__weights': ['uniform', 'distance']
|
||||
#}
|
||||
]
|
||||
]
|
||||
|
||||
gscv_fs = GridSearchCV(pipe
|
||||
, search_space
|
||||
|
|
@ -124,12 +130,6 @@ gscv_fs.fit(X, y)
|
|||
gscv_fs.best_params_
|
||||
gscv_fs.best_score_
|
||||
|
||||
##### CRAP
|
||||
gscv_fs.get_params()['transformers']
|
||||
|
||||
##### END CRAP
|
||||
|
||||
|
||||
# Training best score corresponds to the max of the mean_test<score>
|
||||
train_bscore = round(gscv_fs.best_score_, 2); train_bscore
|
||||
print('\nTraining best score (MCC):', train_bscore)
|
||||
|
|
@ -182,12 +182,19 @@ gscv_fs.best_estimator_.named_steps['fs'].grid_scores_.max()
|
|||
# FS results
|
||||
#============
|
||||
# Now get the features out
|
||||
|
||||
all_features = gscv_fs.feature_names_in_
|
||||
n_all_features = gscv_fs.n_features_in_
|
||||
#all_features = gsfit.feature_names_in_
|
||||
|
||||
sel_features = X.columns[gscv_fs.best_estimator_.named_steps['fs'].get_support()]
|
||||
#sel_features = X.columns[gscv_fs.best_estimator_.named_steps['fs'].get_support()]
|
||||
#n_sf = gscv_fs.best_estimator_.named_steps['fs'].n_features_
|
||||
|
||||
#---------------<<<< HERE
|
||||
#if var_type == 'mixed'
|
||||
sel_features = var_type_colnames[gscv_fs.best_estimator_.named_steps['fs'].get_support()]
|
||||
n_sf = gscv_fs.best_estimator_.named_steps['fs'].n_features_
|
||||
#---------------<<<< HERE
|
||||
|
||||
# get model name
|
||||
model_name = gscv_fs.best_estimator_.named_steps['clf']
|
||||
|
|
@ -218,10 +225,9 @@ print('\nAccuracy on Blind test:', round(accuracy_score(y_bts, bts_predict),2))
|
|||
bts_mcc_score = round(matthews_corrcoef(y_bts, bts_predict),2)
|
||||
|
||||
# Diff b/w train and bts test scores
|
||||
train_test_diff = train_bscore - bts_mcc_score
|
||||
train_test_diff = round(train_bscore - bts_mcc_score,2)
|
||||
print('\nDiff b/w train and blind test score (MCC):', train_test_diff)
|
||||
|
||||
|
||||
# create a dict with all scores
|
||||
lr_btsD = {#'best_model': list(gscv_lr_fit_be_mod.items())
|
||||
#'bts_mcc':None
|
||||
|
|
|
|||
Loading…
Add table
Add a link
Reference in a new issue