Advanced Analytical Theory and Methods: Time Series Analysis - Data Science and Big Data Analytics - page 348

Figure 8.19 Plot of residuals from the fitted (0,1,1) × (1,0,0) 12 model
Figure 8.20 Histogram of the residuals from the fitted (0,1,1) × (1,0,0) 12 model
Figure 8.21 Q-Q plot of the residuals from the fitted (0,1,1) (1,0,0) 12 model
If the normality or the constant variance assumptions do not appear to be true, it
may be necessary to transform the time series prior to fitting the ARIMA model. A
common transformation is to apply a logarithm function.