# scikit-sample.py
# ================
#
# Example of using a ML model with scikit-learn.
#
# Requirements:
#
#  - matplotlib
#  - scikit-learn
#  - pandas

from sklearn.datasets import fetch_california_housing
from sklearn.neighbors import KNeighborsRegressor
from sklearn.preprocessing import StandardScaler
from sklearn.pipeline import Pipeline
from sklearn.model_selection import GridSearchCV
import matplotlib.pylab as plt
import pandas as pd

# Sample dataset
X, y = fetch_california_housing(return_X_y=True)

# We create a pipeline where we scale the data and specify a model
pipe = Pipeline([
    ("scale", StandardScaler()),
    ("model", KNeighborsRegressor(n_neighbors=1))
])

# Get all the pipeline settings
# print(pipe.get_params())

# We can try different hyperparameters on the pipeline via the GridSearch
mod = GridSearchCV(estimator=pipe,
                   param_grid={'model__n_neighbors': [1, 2, 3, 4, 5, 6]},
                   cv=3)

mod.fit(X, y)

df = pd.DataFrame(mod.cv_results_)
plt.plot([i for i in range(6)],
         df['mean_test_score'])
plt.xlabel('n_neighbors')
plt.ylabel('mean_tests_score')
plt.show()