raise MissingDataError('exog contains inf or nans')

import pandas as pd
import matplotlib.pyplot as plt
from statsmodels.tsa.holtwinters import ExponentialSmoothing as HWES
from statsmodels.tsa.stattools import adfuller
from statsmodels.tsa.seasonal import seasonal_decompose
import warnings
 
warnings.simplefilter('ignore')
 
# Load the data
csv_path = r'C:\Users\test\HW_in.csv'
df = pd.read_csv(csv_path, delimiter=',', header=0, infer_datetime_format=True, parse_dates=[0], index_col=[0])
 
# Ensure the index is in datetime format
df.index = pd.to_datetime(df.index)
 
# Set the frequency to monthly
df = df.asfreq('M')
 
# Print the DataFrame and its columns for debugging
print("DataFrame head:\n", df.head())
print("DataFrame columns:\n", df.columns)
 
# Check if 'Suma' is in the columns
if 'Suma' not in df.columns:
    print("Error: 'Suma' column not found in the DataFrame")
else:
    # Perform adfuller test on 'Suma'
    results = adfuller(df['Suma'])
    print('adfuller results raw data:', results)
 
    # Perform differencing and adfuller test on the differenced data
    HW_diff = df.diff().dropna()
    result_diff = adfuller(HW_diff['Suma'])
    print('adfuller results first difference:', result_diff)
 
    # Perform seasonal decomposition
    decomp = seasonal_decompose(df['Suma'], model='additive')
    # decomp.plot()
    # plt.show()
 
    '''
    # Forecast non diff data
    model = HWES(df, seasonal_periods=12, trend='add', seasonal='add')
    fitted = model.fit()
    print(fitted.summary())
    HW_forecast = fitted.forecast(steps=12)
    print(HW_forecast)
    '''
 
    # Forecast diff data
    model = HWES(HW_diff, seasonal_periods=12, trend='add', seasonal='add')
    fitted = model.fit()
    print(fitted.summary())
    HW_forecast = fitted.forecast(steps=16)
    print(HW_forecast)

df.index=df.index +pd.offsets.MonthEnd(0)