tanu/LSHTM_analysis
1
0
Fork 0
Code Issues Pull requests Releases Wiki Activity

added Mult_clfs_logo and Mult_clsf.py with consistency

Browse source
This commit is contained in:
Tanushree Tunstall 2022-07-10 12:32:52 +01:00
parent 06f2ce97b6
commit de5c1270be
5 changed files with 201 additions and 115 deletions
Download patch file Download diff file Expand all files Collapse all files

150
scripts/ml/ml_functions/MultClfs.py
View file

@ -74,10 +74,13 @@ from sklearn.impute import KNNImputer as KNN
import json
import argparse
import re
import itertools
from sklearn.model_selection import LeaveOneGroupOut
from sklearn.decomposition import PCA
#%% GLOBALS
rs = {'random_state': 42}
njobs = {'n_jobs': os.cpu_count() } # the number of jobs should equal the number of CPU cores
#rs = {'random_state': 42}
#njobs = {'n_jobs': os.cpu_count() } # the number of jobs should equal the number of CPU cores
scoring_fn = ({ 'mcc' : make_scorer(matthews_corrcoef)
, 'fscore' : make_scorer(f1_score)
@ -88,13 +91,13 @@ scoring_fn = ({ 'mcc' : make_scorer(matthews_corrcoef)
, 'jcc' : make_scorer(jaccard_score)
})
skf_cv = StratifiedKFold(n_splits = 10
#, shuffle = False, random_state= None)
, shuffle = True,**rs)
#skf_cv = StratifiedKFold(n_splits = 10
# #, shuffle = False, random_state= None)
# , shuffle = True,**rs)
rskf_cv = RepeatedStratifiedKFold(n_splits = 10
, n_repeats = 3
, **rs)
#rskf_cv = RepeatedStratifiedKFold(n_splits = 10
# , n_repeats = 3
# , **rs)
mcc_score_fn = {'mcc': make_scorer(matthews_corrcoef)}
jacc_score_fn = {'jcc': make_scorer(jaccard_score)}
@ -137,6 +140,7 @@ scoreBT_mapD = {'bts_mcc' : 'MCC'
, 'bts_jcc' : 'JCC'
}
#gene_group = 'gene_name'
#%%############################################################################
############################
# MultModelsCl()
@ -145,17 +149,23 @@ scoreBT_mapD = {'bts_mcc' : 'MCC'
# Multiple Classification - Model Pipeline
def MultModelsCl(input_df, target
, sel_cv
, blind_test_df
, blind_test_target
, tts_split_type
, resampling_type = 'none' # default
, tts_split_type
, resampling_type
#, group = None
, add_cm = True # adds confusion matrix based on cross_val_predict
, add_yn = True # adds target var class numbers
, var_type = ['numerical', 'categorical','mixed']
, var_type = ['numerical', 'categorical','mixed']
, scale_numeric = ['min_max', 'std', 'min_max_neg', 'none']
, run_blind_test = True
, return_formatted_output = True):
, blind_test_df = pd.DataFrame()
, blind_test_target = pd.Series(dtype = int)
, return_formatted_output = True
, random_state = 42
, n_jobs = os.cpu_count() # the number of jobs should equal the number of CPU cores
):
'''
@ param input_df: input features
@ -173,7 +183,25 @@ def MultModelsCl(input_df, target
returns
Dict containing multiple classification scores for each model and mean of each Stratified Kfold including training
'''
#%% Func globals
rs = {'random_state': random_state}
njobs = {'n_jobs': n_jobs}
skf_cv = StratifiedKFold(n_splits = 10
#, shuffle = False, random_state= None)
, shuffle = True,**rs)
rskf_cv = RepeatedStratifiedKFold(n_splits = 10
, n_repeats = 3
, **rs)
logo = LeaveOneGroupOut()
# select CV type:
# if group == None:
# sel_cv = skf_cv
# else:
# sel_cv = logo
#======================================================
# Determine categorical and numerical features
#======================================================
@ -196,8 +224,9 @@ def MultModelsCl(input_df, target
# # t = [('num', MinMaxScaler(), numerical_ix)
# # , ('cat', OneHotEncoder(), categorical_ix) ]
# if var_type == 'mixed':
# t = [('cat', OneHotEncoder(), categorical_ix) ]
# col_transform = ColumnTransformer(transformers = t
# , remainder='passthrough')
if type(var_type) == list:
var_type = str(var_type[0])
else:
@ -229,37 +258,37 @@ def MultModelsCl(input_df, target
#======================================================
# Specify multiple Classification Models
#======================================================
models = [('AdaBoost Classifier' , AdaBoostClassifier(**rs) )
# , ('Bagging Classifier' , BaggingClassifier(**rs, **njobs, bootstrap = True, oob_score = True, verbose = 3, n_estimators = 100) )
# , ('Decision Tree' , DecisionTreeClassifier(**rs) )
# , ('Extra Tree' , ExtraTreeClassifier(**rs) )
# , ('Extra Trees' , ExtraTreesClassifier(**rs) )
# , ('Gradient Boosting' , GradientBoostingClassifier(**rs) )
# , ('Gaussian NB' , GaussianNB() )
# , ('Gaussian Process' , GaussianProcessClassifier(**rs) )
# , ('K-Nearest Neighbors' , KNeighborsClassifier() )
, ('LDA' , LinearDiscriminantAnalysis() )
# , ('Logistic Regression' , LogisticRegression(**rs) )
# , ('Logistic RegressionCV' , LogisticRegressionCV(cv = 3, **rs))
# , ('MLP' , MLPClassifier(max_iter = 500, **rs) )
#, ('Multinomial' , MultinomialNB() )
# , ('Naive Bayes' , BernoulliNB() )
# , ('Passive Aggresive' , PassiveAggressiveClassifier(**rs, **njobs) )
# , ('QDA' , QuadraticDiscriminantAnalysis() )
# , ('Random Forest' , RandomForestClassifier(**rs, n_estimators = 1000, **njobs ) )
# # , ('Random Forest2' , RandomForestClassifier(min_samples_leaf = 5
# , n_estimators = 1000
# , bootstrap = True
# , oob_score = True
# , **njobs
# , **rs
# , max_features = 'auto') )
# , ('Ridge Classifier' , RidgeClassifier(**rs) )
# , ('Ridge ClassifierCV' , RidgeClassifierCV(cv = 3) )
# , ('SVC' , SVC(**rs) )
# , ('Stochastic GDescent' , SGDClassifier(**rs, **njobs) )
# , ('XGBoost' , XGBClassifier(**rs, verbosity = 0, use_label_encoder =False, **njobs) )
#
models = [('AdaBoost Classifier' , AdaBoostClassifier(**rs) )
, ('Bagging Classifier' , BaggingClassifier(**rs, **njobs, bootstrap = True, oob_score = True, verbose = 3, n_estimators = 100) )
, ('Decision Tree' , DecisionTreeClassifier(**rs) )
, ('Extra Tree' , ExtraTreeClassifier(**rs) )
, ('Extra Trees' , ExtraTreesClassifier(**rs) )
, ('Gradient Boosting' , GradientBoostingClassifier(**rs) )
, ('Gaussian NB' , GaussianNB() )
, ('Gaussian Process' , GaussianProcessClassifier(**rs) )
, ('K-Nearest Neighbors' , KNeighborsClassifier() )
, ('LDA' , LinearDiscriminantAnalysis() )
, ('Logistic Regression' , LogisticRegression(**rs) )
, ('Logistic RegressionCV' , LogisticRegressionCV(cv = 3, **rs))
, ('MLP' , MLPClassifier(max_iter = 500, **rs) )
, ('Multinomial' , MultinomialNB() )
, ('Naive Bayes' , BernoulliNB() )
, ('Passive Aggresive' , PassiveAggressiveClassifier(**rs, **njobs) )
, ('QDA' , QuadraticDiscriminantAnalysis() )
, ('Random Forest' , RandomForestClassifier(**rs, n_estimators = 1000, **njobs ) )
, ('Random Forest2' , RandomForestClassifier(min_samples_leaf = 5
, n_estimators = 1000
, bootstrap = True
, oob_score = True
, **njobs
, **rs
, max_features = 'auto') )
, ('Ridge Classifier' , RidgeClassifier(**rs) )
, ('Ridge ClassifierCV' , RidgeClassifierCV(cv = 3) )
, ('SVC' , SVC(**rs) )
, ('Stochastic GDescent' , SGDClassifier(**rs, **njobs) )
, ('XGBoost' , XGBClassifier(**rs, verbosity = 0, use_label_encoder = False, **njobs) )
]
mm_skf_scoresD = {}
@ -289,10 +318,11 @@ def MultModelsCl(input_df, target
print('\nRunning model pipeline:', model_pipeline)
skf_cv_modD = cross_validate(model_pipeline
cv_modD = cross_validate(model_pipeline
, input_df
, target
, cv = sel_cv
#, groups = group
, scoring = scoring_fn
, return_train_score = True)
#==============================
@ -300,7 +330,7 @@ def MultModelsCl(input_df, target
#==============================
mm_skf_scoresD[model_name] = {}
for key, value in skf_cv_modD.items():
for key, value in cv_modD.items():
print('\nkey:', key, '\nvalue:', value)
print('\nmean value:', np.mean(value))
mm_skf_scoresD[model_name][key] = round(np.mean(value),2)
@ -308,7 +338,7 @@ def MultModelsCl(input_df, target
# ADD more info: meta data related to input df
mm_skf_scoresD[model_name]['resampling'] = resampling_type
mm_skf_scoresD[model_name]['n_training_size'] = len(input_df)
#mm_skf_scoresD[model_name]['n_trainingY_ratio'] = round(Counter(target)[0]/Counter(target)[1], 2)
mm_skf_scoresD[model_name]['n_trainingY_ratio'] = round(Counter(target)[0]/Counter(target)[1], 2)
mm_skf_scoresD[model_name]['n_features'] = len(input_df.columns)
mm_skf_scoresD[model_name]['tts_split'] = tts_split_type
@ -321,7 +351,12 @@ def MultModelsCl(input_df, target
cmD = {}
# Calculate cm
y_pred = cross_val_predict(model_pipeline, input_df, target, cv = sel_cv, **njobs)
y_pred = cross_val_predict(model_pipeline
, input_df
, target
, cv = sel_cv
#, groups = group
, **njobs)
#_tn, _fp, _fn, _tp = confusion_matrix(y_pred, y).ravel() # internally
tn, fp, fn, tp = confusion_matrix(y_pred, target).ravel()
@ -357,7 +392,7 @@ def MultModelsCl(input_df, target
# Build bts numbers dict
btD = {'n_blindY_neg' : Counter(blind_test_target)[0]
, 'n_blindY_pos' : Counter(blind_test_target)[1]
#, 'n_testY_ratio' : round(Counter(blind_test_target)[0]/Counter(blind_test_target)[1], 2)
, 'n_testY_ratio' : round(Counter(blind_test_target)[0]/Counter(blind_test_target)[1], 2)
, 'n_test_size' : len(blind_test_df) }
# Update cmD+tnD dicts with btD
@ -371,9 +406,9 @@ def MultModelsCl(input_df, target
bts_predict = model_pipeline.predict(blind_test_df)
bts_mcc_score = round(matthews_corrcoef(blind_test_target, bts_predict),2)
print('\nMCC on Blind test:' , bts_mcc_score)
print('\nMCC on Blind test:' , bts_mcc_score)
#print('\nAccuracy on Blind test:', round(accuracy_score(blind_test_target, bts_predict),2))
print('\nMCC on Training:' , mm_skf_scoresD[model_name]['test_mcc'] )
print('\nMCC on Training:' , mm_skf_scoresD[model_name]['test_mcc'] )
mm_skf_scoresD[model_name]['bts_mcc'] = bts_mcc_score
mm_skf_scoresD[model_name]['bts_fscore'] = round(f1_score(blind_test_target, bts_predict),2)
@ -384,7 +419,7 @@ def MultModelsCl(input_df, target
mm_skf_scoresD[model_name]['bts_jcc'] = round(jaccard_score(blind_test_target, bts_predict),2)
#mm_skf_scoresD[model_name]['diff_mcc'] = train_test_diff_MCC
#return(mm_skf_scoresD)
#============================
# Process the dict to have WF
@ -526,7 +561,8 @@ def ProcessMultModelsCl(inputD = {}, blind_test_data = True):
sys.exit('\nFAIL: Could not merge metadata with CV and BT dfs')
else:
print('\nConcatenting dfs not possible [WF],check numbers ')
# print('\nConcatenting dfs not possible [WF],check numbers ')
print('\nOnly combining CV and metadata')
#-------------------------------------
# Combine WF+Metadata: Final output

152
scripts/ml/ml_functions/MultClfs_logo_skf.py
View file

@ -76,7 +76,12 @@ import argparse
import re
import itertools
from sklearn.model_selection import LeaveOneGroupOut
from sklearn.decomposition import PCA
#%% GLOBALS
#rs = {'random_state': 42}
#njobs = {'n_jobs': os.cpu_count() } # the number of jobs should equal the number of CPU cores
scoring_fn = ({ 'mcc' : make_scorer(matthews_corrcoef)
, 'fscore' : make_scorer(f1_score)
, 'precision' : make_scorer(precision_score)
@ -86,7 +91,13 @@ scoring_fn = ({ 'mcc' : make_scorer(matthews_corrcoef)
, 'jcc' : make_scorer(jaccard_score)
})
#skf_cv = StratifiedKFold(n_splits = 10
# #, shuffle = False, random_state= None)
# , shuffle = True,**rs)
#rskf_cv = RepeatedStratifiedKFold(n_splits = 10
# , n_repeats = 3
# , **rs)
mcc_score_fn = {'mcc': make_scorer(matthews_corrcoef)}
jacc_score_fn = {'jcc': make_scorer(jaccard_score)}
@ -139,21 +150,23 @@ scoreBT_mapD = {'bts_mcc' : 'MCC'
def MultModelsCl_logo_skf(input_df
, target
, sel_cv
, blind_test_df = pd.DataFrame()
, blind_test_target = pd.Series(dtype = int)
, tts_split_type = "none"
#, group = 'none'
, resampling_type = 'none' # default
, tts_split_type
, resampling_type
#, group = None
, add_cm = True # adds confusion matrix based on cross_val_predict
, add_yn = True # adds target var class numbers
, var_type = ['numerical', 'categorical','mixed']
, scale_numeric = ['min_max', 'std', 'min_max_neg', 'none']
, run_blind_test = True
, blind_test_df = pd.DataFrame()
, blind_test_target = pd.Series(dtype = int)
, return_formatted_output = True
, random_state = 42
, n_jobs = os.cpu_count() # the number of jobs should equal the number of CPU cores
, ):
):
'''
@ param input_df: input features
@ -165,7 +178,7 @@ def MultModelsCl_logo_skf(input_df
@param skv_cv: stratifiedK fold int or object to allow shuffle and random state to pass
@type: int or StratifiedKfold()
@var_type: numerical, categorical and mixed to determine what col_transform to apply (MinMaxScalar and/or one-ho t encoder)
@var_type: numerical, categorical and mixed to determine what col_transform to apply (MinMaxScalar and/or one-hot encoder)
@type: list
returns
@ -185,8 +198,8 @@ def MultModelsCl_logo_skf(input_df
, **rs)
logo = LeaveOneGroupOut()
# # select CV type:
# if group == 'none':
# select CV type:
# if group == None:
# sel_cv = skf_cv
# else:
# sel_cv = logo
@ -201,52 +214,81 @@ def MultModelsCl_logo_skf(input_df
#======================================================
# Determine preprocessing steps ~ var_type
#======================================================
if var_type == 'numerical':
t = [('num', MinMaxScaler(), numerical_ix)]
# if var_type == 'numerical':
# t = [('num', MinMaxScaler(), numerical_ix)]
# if var_type == 'categorical':
# t = [('cat', OneHotEncoder(), categorical_ix)]
# # if var_type == 'mixed':
# # t = [('num', MinMaxScaler(), numerical_ix)
# # , ('cat', OneHotEncoder(), categorical_ix) ]
# col_transform = ColumnTransformer(transformers = t
# , remainder='passthrough')
if type(var_type) == list:
var_type = str(var_type[0])
else:
var_type = var_type
if var_type in ['numerical','mixed']:
if scale_numeric == ['none']:
t = [('cat', OneHotEncoder(), categorical_ix)]
if scale_numeric != ['none']:
if scale_numeric == ['min_max']:
scaler = MinMaxScaler()
if scale_numeric == ['min_max_neg']:
scaler = MinMaxScaler(feature_range=(-1, 1))
if scale_numeric == ['std']:
scaler = StandardScaler()
t = [('num', scaler, numerical_ix)
, ('cat', OneHotEncoder(), categorical_ix)]
if var_type == 'categorical':
t = [('cat', OneHotEncoder(), categorical_ix)]
if var_type == 'mixed':
t = [('num', MinMaxScaler(), numerical_ix)
, ('cat', OneHotEncoder(), categorical_ix) ]
col_transform = ColumnTransformer(transformers = t
, remainder='passthrough')
#======================================================
# Specify multiple Classification Models
#======================================================
models = [('AdaBoost Classifier' , AdaBoostClassifier(**rs) )
, ('Bagging Classifier' , BaggingClassifier(**rs, **njobs, bootstrap = True, oob_score = True) )
, ('Decision Tree' , DecisionTreeClassifier(**rs) )
, ('Extra Tree' , ExtraTreeClassifier(**rs) )
, ('Extra Trees' , ExtraTreesClassifier(**rs) )
, ('Gradient Boosting' , GradientBoostingClassifier(**rs) )
, ('Gaussian NB' , GaussianNB() )
, ('Gaussian Process' , GaussianProcessClassifier(**rs) )
, ('K-Nearest Neighbors' , KNeighborsClassifier() )
, ('LDA' , LinearDiscriminantAnalysis() )
, ('Logistic Regression' , LogisticRegression(**rs) )
, ('Logistic RegressionCV' , LogisticRegressionCV(cv = 3, **rs))
, ('MLP' , MLPClassifier(max_iter = 500, **rs) )
, ('Multinomial' , MultinomialNB() )
, ('Naive Bayes' , BernoulliNB() )
, ('Passive Aggresive' , PassiveAggressiveClassifier(**rs, **njobs) )
, ('QDA' , QuadraticDiscriminantAnalysis() )
, ('Random Forest' , RandomForestClassifier(**rs, n_estimators = 1000 ) )
, ('Random Forest2' , RandomForestClassifier(min_samples_leaf = 5
, n_estimators = 1000
, bootstrap = True
, oob_score = True
, **njobs
, **rs
, max_features = 'auto') )
, ('Ridge Classifier' , RidgeClassifier(**rs) )
, ('Ridge ClassifierCV' , RidgeClassifierCV(cv = 3) )
, ('SVC' , SVC(**rs) )
, ('Stochastic GDescent' , SGDClassifier(**rs, **njobs) )
, ('XGBoost' , XGBClassifier(**rs, verbosity = 0, use_label_encoder =False) )
models = [('AdaBoost Classifier' , AdaBoostClassifier(**rs) )
, ('Bagging Classifier' , BaggingClassifier(**rs, **njobs, bootstrap = True, oob_score = True, verbose = 3, n_estimators = 100) )
# , ('Decision Tree' , DecisionTreeClassifier(**rs) )
# , ('Extra Tree' , ExtraTreeClassifier(**rs) )
# , ('Extra Trees' , ExtraTreesClassifier(**rs) )
# , ('Gradient Boosting' , GradientBoostingClassifier(**rs) )
# , ('Gaussian NB' , GaussianNB() )
# , ('Gaussian Process' , GaussianProcessClassifier(**rs) )
# , ('K-Nearest Neighbors' , KNeighborsClassifier() )
# , ('LDA' , LinearDiscriminantAnalysis() )
# , ('Logistic Regression' , LogisticRegression(**rs) )
# , ('Logistic RegressionCV' , LogisticRegressionCV(cv = 3, **rs))
# , ('MLP' , MLPClassifier(max_iter = 500, **rs) )
# , ('Multinomial' , MultinomialNB() )
# , ('Naive Bayes' , BernoulliNB() )
# , ('Passive Aggresive' , PassiveAggressiveClassifier(**rs, **njobs) )
# , ('QDA' , QuadraticDiscriminantAnalysis() )
# , ('Random Forest' , RandomForestClassifier(**rs, n_estimators = 1000, **njobs ) )
# , ('Random Forest2' , RandomForestClassifier(min_samples_leaf = 5
# , n_estimators = 1000
# , bootstrap = True
# , oob_score = True
# , **njobs
# , **rs
# , max_features = 'auto') )
# , ('Ridge Classifier' , RidgeClassifier(**rs) )
# , ('Ridge ClassifierCV' , RidgeClassifierCV(cv = 3) )
# , ('SVC' , SVC(**rs) )
# , ('Stochastic GDescent' , SGDClassifier(**rs, **njobs) )
# , ('XGBoost' , XGBClassifier(**rs, verbosity = 0, use_label_encoder = False, **njobs) )
]
mm_skf_scoresD = {}
@ -268,6 +310,12 @@ def MultModelsCl_logo_skf(input_df
model_pipeline = Pipeline([
('prep' , col_transform)
, ('model' , model_fn)])
# model_pipeline = Pipeline([
# ('prep' , col_transform)
# , ('pca' , PCA(n_components = 2))
# , ('model' , model_fn)])
print('\nRunning model pipeline:', model_pipeline)
cv_modD = cross_validate(model_pipeline
@ -358,9 +406,10 @@ def MultModelsCl_logo_skf(input_df
bts_predict = model_pipeline.predict(blind_test_df)
bts_mcc_score = round(matthews_corrcoef(blind_test_target, bts_predict),2)
print('\nMCC on Blind test:' , bts_mcc_score)
print('\nAccuracy on Blind test:', round(accuracy_score(blind_test_target, bts_predict),2))
print('\nMCC on Blind test:' , bts_mcc_score)
#print('\nAccuracy on Blind test:', round(accuracy_score(blind_test_target, bts_predict),2))
print('\nMCC on Training:' , mm_skf_scoresD[model_name]['test_mcc'] )
mm_skf_scoresD[model_name]['bts_mcc'] = bts_mcc_score
mm_skf_scoresD[model_name]['bts_fscore'] = round(f1_score(blind_test_target, bts_predict),2)
mm_skf_scoresD[model_name]['bts_precision'] = round(precision_score(blind_test_target, bts_predict),2)
@ -387,8 +436,7 @@ def MultModelsCl_logo_skf(input_df
############################
#Processes the dict from above if use_formatted_output = True
def ProcessMultModelsCl(inputD = {}
, blind_test_data = True):
def ProcessMultModelsCl(inputD = {}, blind_test_data = True):
scoresDF = pd.DataFrame(inputD)

4
scripts/ml/ml_functions/test_func_singlegene.py
View file

@ -26,7 +26,7 @@ skf_cv = StratifiedKFold(n_splits = 10
# , n_repeats = 3
# , **rs)
# param dict for getmldata()
gene_model_paramD = {'data_combined_model' : False
gene_model_paramD = {'data_combined_model' : False
, 'use_or' : False
, 'omit_all_genomic_features': False
, 'write_maskfile' : False
@ -77,7 +77,7 @@ fooD = MultModelsCl(input_df = df2['X_ros']
, blind_test_df = df2['X_bts']
, blind_test_target = df2['y_bts']
, tts_split_type = spl_type
, resampling_type = 'none' # default
, resampling_type = 'XXXX' # default
, var_type = ['mixed']
, scale_numeric = ['min_max']
, return_formatted_output = False