import numpy as np

from sklearn.model_selection import train_test_split

from sklearn import datasets

from sklearn import linear_model

dataset = datasets.load_iris()

dataset.data.shape

dataset.target.shape

holdout_percent = 0.4

X_train, X_test, y_train, y_test = train_test_split(dataset.data, \

dataset.target, test_size = holdout_percent, random_state = 42)

model = linear_model.LogisticRegression(penalty = 'l2', dual = False, tol = 0.0001, \

C = 1.0, fit_intercept = True).fit(X_train, y_train)

from sklearn.model_selection import cross_val_score

model = linear_model.LogisticRegression(penalty = 'l2', dual = False)

scores = cross_val_score(model, dataset.data, dataset.target, cv = 5)

scores.mean(), scores.std() * 2

cross_val_score(model, dataset.data, dataset.target, cv = 5, scoring = 'f1_macro')

###

from sklearn.pipeline import Pipeline

from sklearn.svm import SVC

from sklearn.decomposition import PCA

# 1

Pipeline([('pca', PCA()), ('classifier', SVC())])

# or 2

from sklearn.pipeline import make_pipeline

pipe = make_pipeline(PCA(), SVC())

pipe.steps[1]

pipe.named_steps["svc"]

pipe.named_steps["svc"].decision_function_shape

###

pipe.fit(X_train, y_train)

pipe.score(X_train, y_train)

pipe.score(X_test, y_test)

model.score(X_test, y_test)

###

param_grid = [

{ 'C': [1, 10, 100, 1000], 'kernel': ['linear']},

{'C': [1, 10, 100, 1000], 'gamma':[0.001, 0.0001], 'kernel': ['rbf']},

]

from sklearn.model_selection import GridSearchCV

model = GridSearchCV(SVC(), param_grid, cv = 5, return_train_score = True)

model.fit(X_train, y_train)

import pandas as pd

pd.DataFrame(model.cv_results_)

##

model = GridSearchCV(pipe, dict(svc__C = [0.1, 10, 100]), cv = 5, return_train_score = True)

model.fit(X_train, y_train)