1. data = pd.read_csv()
2. data.head()
3. data.tail()
4. data.info()
5. data.dtypes

• Identify which is the y (predicted output/dependent variable).
• Identify the features X which are relevant to the prediction accuracy rate of y.

1. data.isnull().sum()

• Check for null values in the dataset.

1. Calculate Correlation Coefficients:

• Statistical measure that show how much two variables are related.
• Common: Pearson Correlation Coefficient (r)

corr = df.corr().round(2) # round to make the visualization more aesthetic

8. Create a heatmap to visualize the correlation matrix

import seaborn as sns
import matplotlib.pyplot as plt

# sets the size of the figure
plt.figure(figsize=(10, 8)) # in inches

# use seaborn heatmap() function to create a heatmap using the correlation matrix,
# adding numerical annotations to each cell, 
# set the color map to coolwarm
# specify the format for the annotation text as 2 decimanl placess
sns.heatmap(corr_matrix, annot=True, cmap='coolwarm', fmt=".2f")

# display the plot
plt.show()

1. Visualize categorical variables

plt.figure(figsize = (20, 25))

plt.subplot(5, 2, 1)
plt.gca().set_title('Variable Geography')
sns.countplot(x = 'Geography', palette = 'Set2', data = df)

plt.subplot(5, 2, 2)
plt.gca().set_title('Variable Gender')
sns.countplot(x = 'Gender', palette = 'Set2', data = df)

plt.subplot(5, 2, 3)
plt.gca().set_title('Variable Tenure')
sns.countplot(x = 'Tenure', palette = 'Set2', data = df)

plt.subplot(5, 2, 4)
plt.gca().set_title('Variable NumOfProducts')
sns.countplot(x = 'NumOfProducts', palette = 'Set2', data = df)

plt.subplot(5, 2, 5)
plt.gca().set_title('Variable HasCrCard')
sns.countplot(x = 'HasCrCard', palette = 'Set2', data = df)

plt.subplot(5, 2, 6)
plt.gca().set_title('Variable IsActiveMember')
sns.countplot(x = 'IsActiveMember', palette = 'Set2', data = df)

plt.subplot(5, 2, 7)
plt.gca().set_title('Variable Exited')
sns.countplot(x = 'Exited', palette = 'Set2', data = df)

plt.subplot(5, 2, 8)
plt.gca().set_title('Variable Complain')
sns.countplot(x = 'Complain', palette = 'Set2', data = df)

plt.subplot(5, 2, 9)
plt.gca().set_title('Variable Satisfaction Score')
sns.countplot(x = 'Satisfaction Score', palette = 'Set2', data = df)

plt.subplot(5, 2, 10)
plt.gca().set_title('Variable Card Type')
sns.countplot(x = 'Card Type', palette = 'Set2', data = df)

• set the figure size for the big figure containing all the subplot.
• plot the subplot: plt.subplot(nrows,ncols,index)
• get the current axis and set the title to the variable name: plt.gca().set_title(’Variable name’)