!pip install regressors
Collecting regressors
Downloading regressors-0.0.3.tar.gz (24 kB)
Preparing metadata (setup.py) ... error
error: subprocess-exited-with-error
× python setup.py egg_info did not run successfully.
│ exit code: 1
╰─> [1 lines of output]
error in regressors setup command: use_2to3 is invalid.
[end of output]
note: This error originates from a subprocess, and is likely not a problem with pip.
error: metadata-generation-failed
× Encountered error while generating package metadata.
╰─> See above for output.
note: This is an issue with the package mentioned above, not pip.
hint: See above for details.
WARNING: You are using pip version 22.0.4; however, version 22.1.2 is available.
You should consider upgrading via the '/root/venv/bin/python -m pip install --upgrade pip' command.
import pandas as pd
import seaborn as sns
sns.set(style='darkgrid', context='notebook')
df = pd.read_csv('/work/dataset/insurance.csv')
df
ageint64
18 - 64
sexobject
male50.5%
female49.5%
0
19
female
1
18
male
2
28
male
3
33
male
4
32
male
5
31
female
6
46
female
7
37
female
8
37
male
9
60
female
df.shape
df.charges.hist(bins=40);
df[df.charges>5E4]
ageint64
sexobject
34
28
male
543
54
female
577
31
female
819
33
female
1146
60
male
1230
52
male
1300
45
male
df = df[df.charges<5E4]
# Eliminamos los cargos mayores, el maestro nos aclaro que esto no es
# siempre recomendable sin antes consultar un experto pero por fines didacticos lo hicimos. """
df
ageint64
18 - 64
sexobject
male50.5%
female49.5%
0
19
female
1
18
male
2
28
male
3
33
male
4
32
male
5
31
female
6
46
female
7
37
female
8
37
male
9
60
female
df.bmi.hist(bins=40);
df.age.hist(bins=30);
df[df.age<=20]
ageint64
18 - 20
sexobject
male51.8%
female48.2%
0
19
female
1
18
male
15
19
male
22
18
male
31
18
female
32
19
female
35
19
male
46
18
female
50
18
female
57
18
male
df.smoker.hist();
df.region.hist();
df.sex.hist();
df.children.hist();
import matplotlib.pyplot as plt
sns.pairplot(df, height=2.5)
plt.show()
import numpy as np
numeric_cols = ['age', 'bmi', 'children', 'charges']
cc = np.corrcoef(df[numeric_cols].values.T)
sns.set(font_scale=1.5)
sns.heatmap(cc, annot=True, yticklabels=numeric_cols, xticklabels=numeric_cols)
df = pd.get_dummies(df, columns=['sex', 'smoker', 'region'], drop_first=True)
df.head()
ageint64
bmifloat64
0
19
27.9
1
18
33.77
2
28
33
3
33
22.705
4
32
28.88
from sklearn.preprocessing import StandardScaler
from sklearn.linear_model import LinearRegression
from sklearn.model_selection import train_test_split # Esto va a crear un dataset de prueba y otro de modelo
# Definimos las columnas que seran nuestras x y cuales van a ser nuestras variables predictorias.
X_cols = list(set(df.columns)-set(['charges']))
# Quitamos la columnas de charges, una ventaja
# de usar 'set()' es que eliminamos columnas duplicadas
y_col = ['charges']
X = df[X_cols].values
y = df[y_col].values
X_train, X_test, y_train, y_test = train_test_split(X, y)
# Aqui estamos definiendo los valores que van
# a servir para entrenar el modelo y posteriormente, los valores que se usaran para probarlo.
sc_x = StandardScaler().fit(X)
sc_y = StandardScaler().fit(y)
# Normalizamos
X_train = sc_x.transform(X_train)
X_test = sc_x.transform(X_test)
y_train = sc_y.transform(y_train)
y_test = sc_y.transform(y_test)
# Definimos el modelo
model = LinearRegression()
model.fit(X_train, y_train)
# Hacemos una predicción
y_pred = model.predict(X_test)
y_pred.shape # Predijo 333 datos y solo tiene una preddicon sobre ellos por que solo tenemos una avarible predictoria
# Importamos las metricas que necesitamos
import sklearn.metrics as metrics
mse = metrics.mean_squared_error(y_test, y_pred)
r2 = metrics.r2_score(y_test, y_pred)
print('r2: ', r2.round(4))
print('mse: ', mse.round(4))
r2: 0.7568
mse: 0.2347
from regressors import stats
# Limpiamos los datos del modelo
# Se necesita que los datos de la pendiente y el intercepto tenga un tipo de forma especifica,
# para que sea mas facil a la libreria analizar los resultados
model.intercept_ = model.intercept_[0] # Solo en el valor
model.coef_ = model.coef_.reshape(-1) # Va permitir que la libreria de stast cuando aplique cuando aplique la funcion que
# necesita aplicar utilice estos coeficientes de la manera en que los necesita
Execution error
ModuleNotFoundError: No module named 'regressors'
y_test = y_test.reshape(-1) # Hacemos un reshape para que la librería lo pueda leer de la manera adecuada
print(10*"=",'Sumary',10*'=')
stats.summary(model, X_test, y_test, X_cols)
residuals = np.subtract(y_test, y_pred.reshape(-1)) # Al momento de crear esta substraccion numpy esta creando un nuevo,
# arreglo con una dimension diferente, generalmente se arregla haciendo un reshape de una de las variables a -1.
# Ya se habia hecho reshape a y_test, por eso solo lo hacemos sobre y_pred.
plt.scatter(y_pred, residuals)
plt.axhline(y=0, color='r', linestyle='-')
plt.show()
df2 = df.copy()
df2['age^2'] = df2.age**2
df2['sobrepeso'] = (df2.bmi >= 30).astype(int)
df2['sobrepeso*fumador'] = df2.sobrepeso * df2.smoker_yes
# Definimos las columnas que seran nuestras x y cuales van a ser nuestras variables predictorias.
X_cols = list(set(df2.columns)-set(['charges']))
# Quitamos la columnas de charges, una ventaja
# de usar 'set()' es que eliminamos columnas duplicadas
y_col = ['charges']
X = df2[X_cols].values
y = df2[y_col].values
X_train, X_test, y_train, y_test = train_test_split(X, y)
# Aqui estamos definiendo los valores que van
# a servir para entrenar el modelo y posteriormente, los valores que se usaran para probarlo.
sc_x = StandardScaler().fit(X)
sc_y = StandardScaler().fit(y)
# Normalizamos
X_train = sc_x.transform(X_train)
X_test = sc_x.transform(X_test)
y_train = sc_y.transform(y_train)
y_test = sc_y.transform(y_test)
# Definimos el modelo
model = LinearRegression()
model.fit(X_train, y_train)
# Hacemos una predicción
y_pred = model.predict(X_test)
mse = metrics.mean_squared_error(y_test, y_pred)
r2 = metrics.r2_score(y_test, y_pred)
print('r2: ', r2.round(4))
print('mse: ', mse.round(4))
r2: 0.7829
mse: 0.1831
model.intercept_ = model.intercept_[0]
model.coef_ = model.coef_.reshape(-1)
y_test = y_test.reshape(-1)
print(10*"=",'Sumary',10*'=')
stats.summary(model, X_test, y_test, X_cols)
========== Sumary ==========
Residuals:
Min 1Q Median 3Q Max
-2.0645 0.0492 0.0948 0.1194 1.4434
Coefficients:
Estimate Std. Error t value p value
_intercept -0.010215 0.023808 -0.4291 0.668152
sex_male -0.018591 0.023645 -0.7863 0.432278
children 0.067048 0.023269 2.8814 0.004217
region_southwest -0.042326 0.027516 -1.5383 0.124936
age -0.014036 0.120928 -0.1161 0.907670
sobrepeso -0.023753 0.036649 -0.6481 0.517356
smoker_yes 0.462820 0.031888 14.5140 0.000000
region_northwest -0.011622 0.028807 -0.4035 0.686874
age^2 0.329831 0.119539 2.7592 0.006116
sobrepeso*fumador 0.496414 0.036864 13.4661 0.000000
region_southeast -0.029199 0.028647 -1.0193 0.308820
bmi 0.048472 0.038321 1.2649 0.206801
---
R-squared: 0.78294, Adjusted R-squared: 0.77550
F-statistic: 105.26 on 11 features
residuals = np.subtract(y_test, y_pred.reshape(-1))
plt.scatter(y_pred, residuals)
plt.axhline(y=0, color='r', linestyle='-')
plt.show()
# Definimos las columnas que seran nuestras x y cuales van a ser nuestras variables predictorias.
X_cols = ['children', 'smoker_yes', 'age^2', 'sobrepeso*fumador']
# Quitamos la columnas de charges, una ventaja
# de usar 'set()' es que eliminamos columnas duplicadas
y_col = ['charges']
X = df2[X_cols].values
y = df2[y_col].values
X_train, X_test, y_train, y_test = train_test_split(X, y)
# Aqui estamos definiendo los valores que van
# a servir para entrenar el modelo y posteriormente, los valores que se usaran para probarlo.
sc_x = StandardScaler().fit(X)
sc_y = StandardScaler().fit(y)
# Normalizamos
X_train = sc_x.transform(X_train)
X_test = sc_x.transform(X_test)
y_train = sc_y.transform(y_train)
y_test = sc_y.transform(y_test)
# Definimos el modelo
model = LinearRegression()
model.fit(X_train, y_train)
# Hacemos una predicción
y_pred = model.predict(X_test)
mse = metrics.mean_squared_error(y_test, y_pred)
r2 = metrics.r2_score(y_test, y_pred)
print('r2: ', r2.round(4))
print('mse: ', mse.round(4))
r2: 0.8572
mse: 0.1345
model.intercept_ = model.intercept_[0]
model.coef_ = model.coef_.reshape(-1)
y_test = y_test.reshape(-1)
print(10*"=",'Sumary',10*'=')
stats.summary(model, X_test, y_test, X_cols)
========== Sumary ==========
Residuals:
Min 1Q Median 3Q Max
-2.0422 0.0778 0.1116 0.1526 0.3362
Coefficients:
Estimate Std. Error t value p value
_intercept 0.002586 0.020217 0.1279 0.898278
children 0.068471 0.021048 3.2531 0.001259
smoker_yes 0.455906 0.024179 18.8556 0.000000
age^2 0.324134 0.020062 16.1565 0.000000
sobrepeso*fumador 0.503351 0.025474 19.7597 0.000000
---
R-squared: 0.85724, Adjusted R-squared: 0.85550
F-statistic: 492.40 on 4 features
residuals = np.subtract(y_test, y_pred.reshape(-1))
plt.scatter(y_pred, residuals)
plt.axhline(y=0, color='r', linestyle='-')
plt.show()
# Definimos las columnas que seran nuestras x y cuales van a ser nuestras variables predictorias.
X_cols = ['children', 'smoker_yes', 'age^2', 'sobrepeso*fumador']
# Quitamos la columnas de charges, una ventaja
# de usar 'set()' es que eliminamos columnas duplicadas
y_col = ['charges']
X = df2[X_cols].values
y = df2[y_col].values
X_train, X_test, y_train, y_test = train_test_split(X, y)
# Aqui estamos definiendo los valores que van
# a servir para entrenar el modelo y posteriormente, los valores que se usaran para probarlo.
sc_x = StandardScaler().fit(X)
sc_y = StandardScaler().fit(y)
# Normalizamos
X_train = sc_x.transform(X_train)
X_test = sc_x.transform(X_test)
y_train = sc_y.transform(y_train)
y_test = sc_y.transform(y_test)
# Definimos el modelo
model = LinearRegression(fit_intercept=False)
model.fit(X_train, y_train)
# Hacemos una predicción
y_pred = model.predict(X_test)
mse = metrics.mean_squared_error(y_test, y_pred)
r2 = metrics.r2_score(y_test, y_pred)
print('r2: ', r2.round(4))
print('mse: ', mse.round(4))
r2: 0.8739
mse: 0.1437
model.coef_ = model.coef_.reshape(-1)
y_test = y_test.reshape(-1)
print(10*"=",'Sumary',10*'=')
stats.summary(model, X_test, y_test, X_cols)
========== Sumary ==========
Residuals:
Min 1Q Median 3Q Max
-2.0463 0.0658 0.1084 0.1427 1.5378
Coefficients:
Estimate Std. Error t value p value
_intercept 0.000000 0.020876 0.0000 1.00000
children 0.077702 0.020263 3.8346 0.00015
smoker_yes 0.459328 0.027442 16.7378 0.00000
age^2 0.325122 0.021024 15.4646 0.00000
sobrepeso*fumador 0.497010 0.024987 19.8910 0.00000
---
R-squared: 0.87386, Adjusted R-squared: 0.87232
F-statistic: 568.07 on 4 features
residuals = np.subtract(y_test, y_pred.reshape(-1))
plt.scatter(y_pred, residuals)
plt.axhline(y=0, color='r', linestyle='-')
plt.show()