Time forecasting




So, I followed this. It is a great article, but it doesn’t show how to forecast future :

Later in comments I see the following and I have the same problem:

b) How can we forecast for future time points (say 12 time points ahead).
Can we use followings still ?
predictions_ARIMA_log = pd.Series(ts_log.ix[0], index=ts_log.index)
predictions_ARIMA_log = predictions_ARIMA_log.add(predictions_ARIMA_diff_cumsum,fill_value=0)
ts_log is not available for future points.

Since the predictions we are doing has got the trend and cycles removed, and also the values are in log, how to do the prediction with these values added. I have checked both the predict and forecast methods which the author mention but it doesn’t help as the values returned are in log and they don’t have ts_log for future points.



Hi @ali79,

Suppose you have taken the log of your series before fitting the model, you should use exp to scale back the values. I have shared a link below for a training course on time series. It covers the steps of predicting the model and making submission.

Happy learning!