A total of 3 functions are given to calculate the Centered Moving Average of a time series: Weighted Moving Average, Simple Moving Average and Exponential Moving Average.
A) Weighted Moving Average
-
timeseries = The dataset in a Time Series format.
-
w = A list of weigths. The default values are [0.2, 0.3, 0.5].
-
graph = If True then the original dataset and the moving average curves will be plotted. The default value is True.
-
horizon = Calculates the prediction h steps ahead. The default value is 0.
B) Simple Moving Average
-
timeseries = The dataset in a Time Series format.
-
n = Indicates the considered number of periods to calculate the moving average. The default value is 2.
-
graph = If True then the original dataset and the moving average curves will be plotted. The default value is True.
-
horizon = Calculates the prediction h steps ahead. The default value is 0.
C) Exponential Moving Average
-
timeseries = The dataset in a Time Series format.
-
alpha = Exponential smoothing parameter. The default value is 0.5.
-
graph = If True then the original dataset and the moving average curves will be plotted. The default value is True.
-
horizon = Calculates the prediction h steps ahead. The default value is 0.
-
optimize = If True then the best "alpha" (exponential smoothing parameter) is calculated by brute force. The default value is False.