preparing_for_model.py

# -*- coding: utf-8 -*-
"""Preparing for Model.ipynb

Automatically generated by Colaboratory.

Original file is located at
    https://colab.research.google.com/drive/18zKZ50n37YxPrP-SWokx14AGklLCynXu
"""

import pandas as pd
import numpy as np

from google.colab import files
uploaded=files.upload()

df = pd.read_csv('stratified_train_set (6).csv')
print(df)

for column in df.columns:
    missing_values = df[column].isna().sum()

    if missing_values > 0:
        print(f"Column '{column}' has {missing_values} missing values.")
    else:
        print(f"Column '{column}' has no missing values.")

total_missing_values = df.isna().sum().sum()
if total_missing_values > 0:
    print(f"\nThe DataFrame has a total of {total_missing_values} missing values.")
else:
    print("\nThe DataFrame has no missing values.")

"""# **Handling text attributes**"""

from sklearn.preprocessing import LabelEncoder

label_encoder = LabelEncoder()

# List of categorical columns to encode
categorical_columns = [ 'position', 'team', 'report_primary_injury', 'weather']

# Apply LabelEncoder to each categorical column
for column in categorical_columns:
    df[column] = label_encoder.fit_transform(df[column])

# Display the updated DataFrame with encoded categorical columns
print(df)

"""# **One Hot Encoding**"""

import pandas as pd
from sklearn.preprocessing import  OneHotEncoder

# Now, extract the columns that need one-hot encoding
columns_to_onehot_encode = [ 'season', 'age' , 'week', 'team', 'position', 'report_primary_injury', 'weather']

# Initialize OneHotEncoder
onehot_encoder = OneHotEncoder(sparse=False, drop='first')

# Fit and transform the data using OneHotEncoder
encoded_data = onehot_encoder.fit_transform(df[columns_to_onehot_encode])

# Create a DataFrame with the encoded data
encoded_df = pd.DataFrame(encoded_data, columns=onehot_encoder.get_feature_names_out(columns_to_onehot_encode))

# Drop the original columns from the original DataFrame and concatenate the encoded DataFrame
df = pd.concat([df.drop(columns=columns_to_onehot_encode), encoded_df], axis=1)

# Display the updated DataFrame
print(df)

"""# **Feature scaling**"""

from sklearn.preprocessing import MinMaxScaler
from sklearn.preprocessing import MinMaxScaler


minmax_scaler = MinMaxScaler()

# Fit and transform the entire DataFrame
scaled_data = minmax_scaler.fit_transform(df)

# Create a DataFrame with the scaled data
scaled_df = pd.DataFrame(scaled_data, columns=df.columns)

# Display the updated DataFrame
print(scaled_df.head())

# Assuming df is your preprocessed DataFrame
scaled_df.to_csv('scaled_train_set.csv', index=False)