Preprocessing¶

In [1]:

# Uncomment the next line and run this cell to install sorix
#!pip install 'sorix @ git+https://github.com/Mitchell-Mirano/sorix.git@main'

# Uncomment the next line and run this cell to install sorix #!pip install 'sorix @ git+https://github.com/Mitchell-Mirano/sorix.git@main'

In [2]:

import numpy as np
import pandas as pd
from sorix.preprocessing import OneHotEncoder
from sorix.preprocessing import StandardScaler,MinMaxScaler,RobustScaler
from sorix.preprocessing import ColumnTransformer

import numpy as np import pandas as pd from sorix.preprocessing import OneHotEncoder from sorix.preprocessing import StandardScaler,MinMaxScaler,RobustScaler from sorix.preprocessing import ColumnTransformer

In [3]:

data = {'Edad': [25, 30, 45, 50, 35, 60, 20, 40],
        'Ingresos': [30000, 50000, 100000, 120000, 70000, 150000, 20000, 80000],
        'Pais': ['EE. UU.', 'Canadá', 'México', 'EE. UU.', 'Canadá', 'México', 'EE. UU.', 'Canadá'],
        'Ciudades': ['New York', 'Toronto', 'Mexico City', 'New York', 'Toronto', 'Mexico City', 'New York', 'Toronto'],
        'Compra': [0, 1, 1, 1, 0, 1, 0, 1]}
df = pd.DataFrame(data)

X = df.drop('Compra', axis=1)
y = df['Compra']
X

data = {'Edad': [25, 30, 45, 50, 35, 60, 20, 40], 'Ingresos': [30000, 50000, 100000, 120000, 70000, 150000, 20000, 80000], 'Pais': ['EE. UU.', 'Canadá', 'México', 'EE. UU.', 'Canadá', 'México', 'EE. UU.', 'Canadá'], 'Ciudades': ['New York', 'Toronto', 'Mexico City', 'New York', 'Toronto', 'Mexico City', 'New York', 'Toronto'], 'Compra': [0, 1, 1, 1, 0, 1, 0, 1]} df = pd.DataFrame(data) X = df.drop('Compra', axis=1) y = df['Compra'] X

Out[3]:

	Edad	Ingresos	Pais	Ciudades
0	25	30000	EE. UU.	New York
1	30	50000	Canadá	Toronto
2	45	100000	México	Mexico City
3	50	120000	EE. UU.	New York
4	35	70000	Canadá	Toronto
5	60	150000	México	Mexico City
6	20	20000	EE. UU.	New York
7	40	80000	Canadá	Toronto

In [4]:

X.shape

X.shape

Out[4]:

(8, 4)

In [5]:

categorical_features = ['Pais', 'Ciudades']
numeric_features = ['Edad', 'Ingresos'] 

categorical_features = ['Pais', 'Ciudades'] numeric_features = ['Edad', 'Ingresos']

One Hot Encoder¶

In [6]:

encoder = OneHotEncoder()
X_encoded = encoder.fit_transform(X[categorical_features])
pd.DataFrame(X_encoded, columns=encoder.get_features_names())

encoder = OneHotEncoder() X_encoded = encoder.fit_transform(X[categorical_features]) pd.DataFrame(X_encoded, columns=encoder.get_features_names())

Out[6]:

	Pais_Canadá	Pais_EE. UU.	Pais_México	Ciudades_Mexico City	Ciudades_New York	Ciudades_Toronto
0	0.0	1.0	0.0	0.0	1.0	0.0
1	1.0	0.0	0.0	0.0	0.0	1.0
2	0.0	0.0	1.0	1.0	0.0	0.0
3	0.0	1.0	0.0	0.0	1.0	0.0
4	1.0	0.0	0.0	0.0	0.0	1.0
5	0.0	0.0	1.0	1.0	0.0	0.0
6	0.0	1.0	0.0	0.0	1.0	0.0
7	1.0	0.0	0.0	0.0	0.0	1.0

Standard Scaler¶

In [7]:

scaler = StandardScaler() 
X_scaled = scaler.fit_transform(X[numeric_features]) 
pd.DataFrame(X_scaled, columns=scaler.get_features_names())

scaler = StandardScaler() X_scaled = scaler.fit_transform(X[numeric_features]) pd.DataFrame(X_scaled, columns=scaler.get_features_names())

Out[7]:

	Edad	Ingresos
0	-1.051315	-1.137500
1	-0.650814	-0.658553
2	0.550689	0.538816
3	0.951190	1.017763
4	-0.250313	-0.179605
5	1.752192	1.736185
6	-1.451816	-1.376974
7	0.150188	0.059868

One Hot Ecoder + Standard Scaler¶

In [8]:

encoder = OneHotEncoder()
scaler = StandardScaler() 
X_encoded = encoder.fit_transform(X[categorical_features])
X_scaled = scaler.fit_transform(X[numeric_features]) 
X_train = np.hstack((X_encoded, X_scaled))
pd.DataFrame(X_train, columns=encoder.get_features_names() + scaler.get_features_names())

encoder = OneHotEncoder() scaler = StandardScaler() X_encoded = encoder.fit_transform(X[categorical_features]) X_scaled = scaler.fit_transform(X[numeric_features]) X_train = np.hstack((X_encoded, X_scaled)) pd.DataFrame(X_train, columns=encoder.get_features_names() + scaler.get_features_names())

Out[8]:

	Pais_Canadá	Pais_EE. UU.	Pais_México	Ciudades_Mexico City	Ciudades_New York	Ciudades_Toronto	Edad	Ingresos
0	0.0	1.0	0.0	0.0	1.0	0.0	-1.051315	-1.137500
1	1.0	0.0	0.0	0.0	0.0	1.0	-0.650814	-0.658553
2	0.0	0.0	1.0	1.0	0.0	0.0	0.550689	0.538816
3	0.0	1.0	0.0	0.0	1.0	0.0	0.951190	1.017763
4	1.0	0.0	0.0	0.0	0.0	1.0	-0.250313	-0.179605
5	0.0	0.0	1.0	1.0	0.0	0.0	1.752192	1.736185
6	0.0	1.0	0.0	0.0	1.0	0.0	-1.451816	-1.376974
7	1.0	0.0	0.0	0.0	0.0	1.0	0.150188	0.059868

One Hot Ecoder + MinMax Scaler¶

In [9]:

encoder = OneHotEncoder()
scaler = MinMaxScaler()
X_encoded = encoder.fit_transform(X[categorical_features])
X_scaled = scaler.fit_transform(X[numeric_features])
X_train = np.hstack((X_encoded, X_scaled))
pd.DataFrame(X_train, columns=encoder.get_features_names() + scaler.get_features_names())

encoder = OneHotEncoder() scaler = MinMaxScaler() X_encoded = encoder.fit_transform(X[categorical_features]) X_scaled = scaler.fit_transform(X[numeric_features]) X_train = np.hstack((X_encoded, X_scaled)) pd.DataFrame(X_train, columns=encoder.get_features_names() + scaler.get_features_names())

Out[9]:

	Pais_Canadá	Pais_EE. UU.	Pais_México	Ciudades_Mexico City	Ciudades_New York	Ciudades_Toronto	Edad	Ingresos
0	0.0	1.0	0.0	0.0	1.0	0.0	0.125	0.076923
1	1.0	0.0	0.0	0.0	0.0	1.0	0.250	0.230769
2	0.0	0.0	1.0	1.0	0.0	0.0	0.625	0.615385
3	0.0	1.0	0.0	0.0	1.0	0.0	0.750	0.769231
4	1.0	0.0	0.0	0.0	0.0	1.0	0.375	0.384615
5	0.0	0.0	1.0	1.0	0.0	0.0	1.000	1.000000
6	0.0	1.0	0.0	0.0	1.0	0.0	0.000	0.000000
7	1.0	0.0	0.0	0.0	0.0	1.0	0.500	0.461538

One Hot Ecoder + Robust Scaler¶

In [10]:

encoder = OneHotEncoder()
scaler = RobustScaler()
X_encoded = encoder.fit_transform(X[categorical_features])
X_scaled = scaler.fit_transform(X[numeric_features])
X_train = np.hstack((X_encoded, X_scaled))
pd.DataFrame(X_train, columns=encoder.get_features_names() + scaler.get_features_names())

encoder = OneHotEncoder() scaler = RobustScaler() X_encoded = encoder.fit_transform(X[categorical_features]) X_scaled = scaler.fit_transform(X[numeric_features]) X_train = np.hstack((X_encoded, X_scaled)) pd.DataFrame(X_train, columns=encoder.get_features_names() + scaler.get_features_names())

Out[10]:

	Pais_Canadá	Pais_EE. UU.	Pais_México	Ciudades_Mexico City	Ciudades_New York	Ciudades_Toronto	Edad	Ingresos
0	0.0	1.0	0.0	0.0	1.0	0.0	-0.714286	-0.750000
1	1.0	0.0	0.0	0.0	0.0	1.0	-0.428571	-0.416667
2	0.0	0.0	1.0	1.0	0.0	0.0	0.428571	0.416667
3	0.0	1.0	0.0	0.0	1.0	0.0	0.714286	0.750000
4	1.0	0.0	0.0	0.0	0.0	1.0	-0.142857	-0.083333
5	0.0	0.0	1.0	1.0	0.0	0.0	1.285714	1.250000
6	0.0	1.0	0.0	0.0	1.0	0.0	-1.000000	-0.916667
7	1.0	0.0	0.0	0.0	0.0	1.0	0.142857	0.083333

Column Transformer¶

In [11]:

column_transformer = ColumnTransformer(
    transformers = [
        ('cat', OneHotEncoder(), categorical_features),
        ('num', StandardScaler(), numeric_features)
    ]
)

X_train = column_transformer.fit_transform(X)

pd.DataFrame(X_train, columns=column_transformer.get_features_names())

column_transformer = ColumnTransformer( transformers = [ ('cat', OneHotEncoder(), categorical_features), ('num', StandardScaler(), numeric_features) ] ) X_train = column_transformer.fit_transform(X) pd.DataFrame(X_train, columns=column_transformer.get_features_names())

Out[11]:

	cat_Pais_Canadá	cat_Pais_EE. UU.	cat_Pais_México	cat_Ciudades_Mexico City	cat_Ciudades_New York	cat_Ciudades_Toronto	num_Edad	num_Ingresos
0	0.0	1.0	0.0	0.0	1.0	0.0	-1.051315	-1.137500
1	1.0	0.0	0.0	0.0	0.0	1.0	-0.650814	-0.658553
2	0.0	0.0	1.0	1.0	0.0	0.0	0.550689	0.538816
3	0.0	1.0	0.0	0.0	1.0	0.0	0.951190	1.017763
4	1.0	0.0	0.0	0.0	0.0	1.0	-0.250313	-0.179605
5	0.0	0.0	1.0	1.0	0.0	0.0	1.752192	1.736185
6	0.0	1.0	0.0	0.0	1.0	0.0	-1.451816	-1.376974
7	1.0	0.0	0.0	0.0	0.0	1.0	0.150188	0.059868