Scribd
Upload
22 views13 pages

Arma Model

Arma auto regressive moving average

Uploaded by

Saikiran Sai
Copyright
© © All Rights Reserved
We take content rights seriously. If you suspect this is your content, claim it here.
Available Formats
Download as PPTX, PDF, TXT or read online on Scribd
22 views13 pages

Arma Model

Arma auto regressive moving average

Uploaded by

Saikiran Sai
Copyright
© © All Rights Reserved
We take content rights seriously. If you suspect this is your content, claim it here.
Available Formats
Download as PPTX, PDF, TXT or read online on Scribd
You are on page 1/ 13

ARMA MODEL

• Autoregressive Moving Average (ARMA)


models are used for modeling time series
data.
• They combine two components: an
autoregressive (AR) part and a moving average
(MA) part.
Stationarity and Invertibility of ARMA
Models
• Stationarity
• A time series is stationary if its statistical
properties, such as mean and variance, do not
change over time.
• For an ARMA model to be stationary, the
autoregressive (AR) part of the model must
meet certain conditions.
If all the roots of Eq. are less than one in
absolute value, then ARMA(p, q) is stationary
• Invertibility
• Invertibility is a property of the moving
average (MA) part of the ARMA model.
• An MA process is invertible if it can be
expressed as an infinite AR process, which
ensures that past values and shocks can be
recovered from current and past observations.
If all the roots of Eq. are less than one in
absolute value, then ARMA(p, q) is said to be
invertible and has an infinite AR
representation,
• Checking Stationarity
• Stationarity can be checked using:
• Plots: Visual inspection of the time series plot.
• Summary Statistics: Calculating rolling
statistics (mean, variance) over time.
• Statistical Tests: Using tests like the
Augmented Dickey-Fuller (ADF) test.
An increasing variogram suggests that the
time series is non-stationary.
• import numpy as np
• import pandas as pd
• import matplotlib.pyplot as plt

• # Function to compute the variogram


• def compute_variogram(series, max_lag):
• n = len(series)
• variogram = []
• for lag in range(1, max_lag + 1):
• diff = (series[lag:] - series[:-lag]) ** 2
• variogram.append(0.5 * np.mean(diff))
• return variogram
Checking for Stationarity using Variogram

• # Generate a stationary time series (white noise)


• np.random.seed(42)
• n = 100
• stationary_series = np.random.normal(0, 1, n)

• # Compute the variogram for the stationary series


• max_lag = 20
• variogram_stationary =
compute_variogram(stationary_series, max_lag)
A flat variogram suggests that the time
series is stationary.
• # Plot the variogram for the stationary series
• plt.figure(figsize=(10, 5))
• plt.plot(range(1, max_lag + 1), variogram_stationary,
marker='o')
• plt.title('Variogram of Stationary Time Series (White
Noise)')
• plt.xlabel('Lag')
• plt.ylabel('Variogram')
• plt.grid(True)
• plt.show()

You might also like