covres.py

import os
import pandas as pd
import sklearn as skl
import scipy as sp
import matplotlib.pyplot as plt 

# Define  the paths 
BASEPATH = '/home/giulio/Repositories/covres/' #This is the folder where db is

# import database
df = pd.read_csv(BASEPATH + 'Raw/db.csv')

# This generates the heatmap
corr=df.corr() #This gets the correlation of the dataset
plt.figure(figsize=(10,10)) #change figsize to get a smaller or bigger figure
plt.xticks(range(len(df.columns)),df.columns,rotation=90)
plt.yticks(range(len(df.columns)),df.columns)
plt.imshow(corr, cmap='hot',interpolation="nearest")
plt.colorbar() 

#save corr to file
corr.to_csv(BASEPATH + 'Processed/corr.csv')

# This section gets the number of missing values per column

X = []

print("Percentage of Missing values")
for col in df.columns:
    X.append([col, len(df[df[col].isnull()]) / len(df) * 100])
    
missings = pd.DataFrame(X, columns=['Variable', '% Missing'])