manojvenaram / tsa_exp6

Ex.No: 6 HOLT WINTERS METHOD

Date: 30-03-2024

AIM:

To implement the Holt Winters Method Model using Python.

ALGORITHM:

1. You import the necessary libraries
2. You load a CSV file containing daily airpassengers data into a DataFrame, parse the 'date' column as datetime, and perform some initial data exploration
3. You group the data by date and resample it to a monthly frequency (beginning of the month
4. You plot the time series data
5. You import the necessary 'statsmodels' libraries for time series analysis
6. You decompose the time series data into its additive components and plot them:
7. You calculate the root mean squared error (RMSE) to evaluate the model's performance
8. You calculate the mean and standard deviation of the entire sales dataset, then fit a Holt- Winters model to the entire dataset and make future predictions
9. You plot the original passengers data and the predictions

PROGRAM:

Import the libraries

import pandas as pd
import numpy as np
import matplotlib.pyplot as plt
from statsmodels.tsa.holtwinters import ExponentialSmoothing

Load the dataset

data=pd.read_csv("/content/AirPassengers.csv")
data

Convert the string format to Date format

data['Month'] = pd.to_datetime(data['Month'])
data.set_index('Month', inplace=True)

Perform Holt-Winters exponential smoothing

model = ExponentialSmoothing(data, trend="add", seasonal="add", seasonal_periods=12)
fit = model.fit()

Forecast for the next n steps

n_steps = 12  

forecast = fit.forecast(steps=n_steps)
forecast

Plot the original data and the forecast

plt.figure(figsize=(10, 6))
plt.plot(data.index, data, label='Original Data')
plt.plot(pd.date_range(start=data.index[-1], periods=n_steps+1, freq='M')[1:], forecast, label='Forecast')
plt.xlabel('Date')
plt.ylabel('Value')
plt.title('Holt-Winters Forecast')
plt.legend()
plt.show()

OUTPUT:

RESULT:

Thus the program run successfully based on the Holt Winters Method model.