prediction.py

#!/usr/bin/env python3
# -*- coding: utf-8 -*-

import pandas as pd

url = 'https://raw.githubusercontent.com/pcm-dpc/COVID-19/master/dati-andamento-nazionale/dpc-covid19-ita-andamento-nazionale.csv'

usecols=[
         'data', 
         'tamponi', 
         'nuovi_positivi', 
         'deceduti']
df = pd.read_csv(url, usecols=usecols)
df['data'] = pd.to_datetime(df['data'], format="%Y-%m-%dT%H:00:00")
df = df[df['data']>'2020-07-31T17:00:00']

r = range(1, 30)
l = len(df)

df['delta_tamponi'] = 0.0
df['delta_deceduti'] = 0.0
for i in r:
  df['delta_deceduti_%d' % i] = None

df.reset_index(inplace=True)

for idx in df.index:
  if idx>0:
    df.at[idx, 'delta_tamponi'] = df['tamponi'][idx] - df['tamponi'][idx-1]
    df.at[idx, 'delta_deceduti'] = df['deceduti'][idx] - df['deceduti'][idx-1]

for idx in df.index:
  for i in r:
    if idx+i<l:
      df.at[idx, 'delta_deceduti_%d' % i] = df['delta_deceduti'][idx+i]

from sklearn.model_selection import train_test_split
from sklearn.linear_model import LinearRegression

from sklearn.preprocessing import StandardScaler
from sklearn.metrics import r2_score

import numpy as np


ss = StandardScaler()

models = []
scores = []
predictions = []
for i in r:

  target = 'delta_deceduti_%d' % i
  df2 = df.dropna(subset=[target], inplace=False)
  X = df2[['delta_tamponi', 'nuovi_positivi']]
  y = df2[target]
  partial_scores = []
  for j in range(100):
    X_train, X_test, y_train, y_test = train_test_split(X, y, test_size=.2)

    X_train = ss.fit_transform(X_train)
    X_test = ss.transform(X_test)

    model = LinearRegression()
    model.fit(X_train, y_train)
    y_pred = model.predict(X_test)
    partial_scores.append(r2_score(y_pred, y_test))

  scores.append(np.mean(partial_scores))


import matplotlib.pyplot as plt

line = plt.bar(r, scores)
line.set_label('scores')
plt.legend()
plt.gca().set(title='R2 score of the regression models', xlabel='n', ylabel='r2 score')