import pandas as pd
from google.colab import drive
import matplotlib.pyplot as plt
import datetime as dt
import plotly.express as px
from pandas_profiling import ProfileReport




drive.mount('/content/drive')

file = ('/content/drive/MyDrive/00_Energy Data/01_Historical_elec_dem_Swissgrid/EnergieUebersichtCH-2009.xls')
sheet='Zeitreihen0h15'
data = pd.read_excel(file, sheet_name=sheet, index_col=0)

data

data.rename(columns = {'Summe verbrauchte Energie Regelblock Schweiz\nTotal energy consumption Swiss controlblock':'consumption'}, inplace = True)

fig = px.line(data, y="consumption")
fig.show()

data.drop(data.columns.difference(['consumption']), 1, inplace=True)

data

data = data.iloc[1:]
#This drops the frist row

data.describe()

fig = px.violin(data,  y="consumption")
fig.show()

data.reset_index(inplace=True)
data.rename(columns = {'index':'datetime'}, inplace = True)
data

data_hourly = data.groupby([pd.Grouper(key='datetime', freq='h')])['consumption'].sum().reset_index().sort_values('datetime')

fig = px.line(data_hourly, y="consumption")
fig.show()

data_daily = data.groupby([pd.Grouper(key='datetime', freq='D')])['consumption'].sum().reset_index().sort_values('datetime')