Search code examples
pythonpandasmatplotlibstatsmodels

How to get a better view/extract the values from acf_plot

I wrote the following code to get an autocorrelation plot of a time series that I have:

fig, axes = plt.subplots(3, 2, sharex=True)
axes[0, 0].plot(data['Consumptieprijsindex']); axes[0, 0].set_title('Original Series')
plot_acf(data['Consumptieprijsindex'], ax=axes[0, 1])

axes[1, 0].plot(data['Consumptieprijsindex'].diff()); axes[1, 0].set_title('1st Order Differencing')
plot_acf(data['Consumptieprijsindex'].diff().dropna(), ax=axes[1, 1])

axes[2, 0].plot(data['Consumptieprijsindex'].diff().diff()); axes[2, 0].set_title('2nd Order Differencing')
plot_acf(data['Consumptieprijsindex'].diff().diff().dropna(), ax=axes[2, 1])

plt.show()

And the output is the following chart:

enter image description here

Is there any way to zoom in on the initial values of the autocorrelation plot or at least get the values? I need these to be a bit more clear to set the hyperparameters for my SARIMA analysis.

Solution

  • I checked some other fora and apparently you can use the statsmodels acf function to get the values themselves. Furthermore, you can edit the number of shown autocorrelation with the 'lags' parameter in the plot_acf function and the 'nlags' parameter in the acf function.