seaborn

import seaborn as sns

Configuration

plt.figure(figsize=(width,height))

Bar Graph

sns.barplot(data=df, 

            x='column_name',

            y='column_name', 

            hue = 'groupby'

            order=a_list, 

            );

plt.xlabel("xlabel");

plt.ylabel("ylabel");

plt.title('graph title');

Boxplot

sns.boxplot(data=df, 

            x='xcol',

            y='ycol', 

            order=a_list, 

            showfliers=False   #outliers

            );

plt.title('graph title');

Histogram

sns.histplot(data=df['column_name'], 

               binrange=(1,100),

               binwidth=5,

               color = '#FF00FF',

               );

plt.title('graph title');

Heatmap

g=sns.histplot(df, cmap='Blues')

               colorbar=g.collections[0].colorbar

               colorbat.set_ticks([0,1,2,3])

               colorbat.set_ticklabels(['bad','ok','good','excellent'])

plt.show()

Pairplot (Scatterplot Matrix)

sns.pairplot(df)

OLS (Ordinary Least Squares) Functions

from statsmodels.formula.api import ols

OLS = ols(formula = 'y_col ~ x_col', data = df)

OR

OLS = ols(formula = 'y_col ~ x_col + x1_col + C(x2_col)', data = df)

model = OLS.fit()

model.summary()

Regression Line

sns.regplot(x = "x_col", y = "y_col", data = df)

X = df['x_col']

Y_hat = model.predict(X)

residuals = model.resid

sns.scatterplot(x = model.fittedvalues, y = model.resid)

Q-Q Plot (Normality Check)

import matplotlib.pyplot as plt

import statsmodels.api as sm

sm.qqplot(residuals, line = 's')

plt.show()

Homoscedasticity Check

g = sns.scatterplot(x=Y_hat, y=residuals)

g.set_xlabel("Y_hat")

g.set_ylabel("Residuals")

g.axhline(0)

plt.show()

Correlation Heatmap

sns.heatmap(df.corr(method='pearson'), 

            vmin=-1,

            vmax=1,

            annot=True,

            cmap='coolwarm');