I am forecasting solar irradiance using different time series models. Rather than predict the irradiance from $t=0 \text{ min}$ to $t=120 \text{ min}$, I would like to predict the change in irradiance relative to $t=0$ such that for each prediction, the prediction would be $\text{Relative Irradiance}_{t+30} = \text{Irradiance}_{t+30} - \text{Irradiance}_{t}$. Right now, I have all my data in a pandas dataframe.

TLDR: How can I format data for ML prediction of a new relative response variable: the change in the variable from time of prediction.


1 Answer 1


You can just add a "Relative Irradiance" column and predict it instead of "Irradiance"

Translated in python, something like this should do the job

df["Relative_Irradiance"] = df["Irradiance"].values - df["Irradiance"].values[0]

Your Answer

By clicking “Post Your Answer”, you agree to our terms of service and acknowledge you have read our privacy policy.

Not the answer you're looking for? Browse other questions tagged or ask your own question.