DATA SCIENCE PYTHON PLAYGROUND

Future observations must not teach the past.

Chronological slicing, TimeSeriesSplit, mixed preparation and last-block diagnosis.

Revisit: Respect time · Regression checkpoint

Sort by date/hour; reserve the final period; compare forward-validation evidence and diagnose the last training block.

This complete script replaces the editor contents. It includes the input setup and the workflow.

import numpy as np
import pandas as pd
df=pd.read_csv('data/ml-learning/ML-X04.csv')
from sklearn.base import clone
from sklearn.model_selection import train_test_split, TimeSeriesSplit, cross_validate, cross_val_predict, GridSearchCV
from sklearn.dummy import DummyRegressor
from sklearn.metrics import root_mean_squared_error

import matplotlib.pyplot as plt
import seaborn as sns
X=df[['Temperature(°C)', 'Humidity(%)', 'Wind speed (m/s)', 'Visibility (10m)', 'Solar Radiation (MJ/m2)', 'Rainfall(mm)', 'Snowfall (cm)', 'Holiday', 'Hour', 'Seasons']]
y=df['Rented Bike Count']
cut=int(len(df)*.8)
X_train,X_test=X.iloc[:cut],X.iloc[cut:]
y_train,y_test=y.iloc[:cut],y.iloc[cut:]
folds=TimeSeriesSplit(n_splits=5)
candidates={}
grids={}
# Prepare inputs and construct the candidate pipeline.
numeric_features = ['Temperature(°C)', 'Humidity(%)', 'Wind speed (m/s)', 'Visibility (10m)', 'Solar Radiation (MJ/m2)', 'Rainfall(mm)', 'Snowfall (cm)']
categorical_features = ['Hour', 'Seasons']
encoded_binary_features = ['Holiday']
from sklearn.preprocessing import OneHotEncoder
from sklearn.compose import ColumnTransformer
category_encoder = OneHotEncoder(handle_unknown='ignore', sparse_output=False, drop='first')
preprocessor = ColumnTransformer([('continuous', 'passthrough', numeric_features), ('encoded', category_encoder, encoded_binary_features + categorical_features)], verbose_feature_names_out=False)
from sklearn.pipeline import Pipeline
from sklearn.linear_model import LinearRegression
model = LinearRegression()
pipeline = Pipeline([('prepare', preprocessor), ('model', model)])
model = pipeline
candidates['multiple_linear']=model
grids['multiple_linear']=None
# Prepare inputs and construct the candidate pipeline.
numeric_features = ['Temperature(°C)', 'Humidity(%)', 'Wind speed (m/s)', 'Visibility (10m)', 'Solar Radiation (MJ/m2)', 'Rainfall(mm)', 'Snowfall (cm)']
categorical_features = ['Hour', 'Seasons']
encoded_binary_features = ['Holiday']
from sklearn.preprocessing import OneHotEncoder
from sklearn.compose import ColumnTransformer
category_encoder = OneHotEncoder(handle_unknown='ignore', sparse_output=False, drop=None)
preprocessor = ColumnTransformer([('continuous', 'passthrough', numeric_features), ('encoded', category_encoder, encoded_binary_features + categorical_features)], verbose_feature_names_out=False)
from sklearn.pipeline import Pipeline
from sklearn.tree import DecisionTreeRegressor
model = DecisionTreeRegressor(random_state=42)
pipeline = Pipeline([('prepare', preprocessor), ('model', model)])
model = pipeline
candidates['regression_tree']=model
grids['regression_tree']={'model__max_depth': [3, 5, None]}
# Validate initial candidates on matching training folds.
initial_results={name:cross_validate(candidate,X_train,y_train,cv=folds,scoring='neg_root_mean_squared_error') for name,candidate in candidates.items()}
# Compare with a simple reference before tuning.
reference_results=cross_validate(DummyRegressor(strategy='mean'),X_train,y_train,cv=folds,scoring='neg_root_mean_squared_error')
rows=[]
selected={}
for name,candidate in candidates.items():
    initial=initial_results[name]
    grid=grids[name]
    if grid:
        search=GridSearchCV(candidate,grid,cv=folds,scoring='neg_root_mean_squared_error',n_jobs=1)
        search.fit(X_train,y_train)
        selected[name]=search.best_estimator_
        validation_score=float(search.best_score_)
        settings=search.best_params_
    else:
        selected[name]=candidate
        validation_score=float(np.mean(initial['test_score']))
        settings='Keep defaults: no production search'
    rows.append({'model':name,'initial_score':float(np.mean(initial['test_score'])),'selected_score':validation_score,'settings':str(settings)})
cv_results=pd.DataFrame(rows).set_index('model')
# This solution nominates by mean training-fold score. Other evidence-based choices can be defensible.
chosen_name=cv_results.selected_score.idxmax()
chosen=selected[chosen_name]
train_indices,validation_indices=list(folds.split(X_train))[-1]
diagnostic_model=clone(chosen).fit(X_train.iloc[train_indices],y_train.iloc[train_indices])
diagnostic_X=X_train.iloc[validation_indices]
diagnostic_y=y_train.iloc[validation_indices]
oof_predictions=diagnostic_model.predict(diagnostic_X)
residuals=pd.DataFrame({'actual':diagnostic_y,'predicted':oof_predictions,'residual':diagnostic_y-oof_predictions})
fig,ax=plt.subplots(figsize=(6,4))
sns.scatterplot(x=residuals.predicted,y=residuals.residual,s=12,ax=ax)
ax.axhline(0,color='black')
ax.set(xlabel='Training-validation prediction',ylabel='Residual',title='Training-only residuals')
fig.tight_layout()
fig.savefig('diagnostic.png',dpi=150,bbox_inches='tight')
final_model=clone(chosen).fit(X_train,y_train)
final_predictions=final_model.predict(X_test)
final_rmse=root_mean_squared_error(y_test,final_predictions)
print('Final RMSE:',final_rmse)

import numpy as np
import pandas as pd
df=pd.read_csv('data/ml-learning/ML-X04.csv')

Why these steps exist

• The supplied file already fixes the chronological population. The final period stays later than development rows; forward folds and last-block diagnosis preserve that ordering during selection. Weather availability must still be justified for the intended prediction time.

Another valid approach

Equivalent ordinary Python is welcome. A defensible candidate or grouping need not match the solution’s choice; objective evidence must remain consistent.