I have 1 week of data formatted in the following way in R.

$$ \begin{array}{c} \text{Category1} & \text{Category2} & \text{Category3} & \text{Date} & \text{Metric1} & \text{Metric2} & \text{...} & \text{Metric7} \\ \hline X & a & 1 & 7/1/17 \\ X & a & 2 & 7/2/17 \\ X & b & 1 & 7/3/17 \end{array} $$

My goal is to predict Metric1 on a future date, ex. 7/15/17.

1st question - can you roughly describe how you might approach this? I have little time-series experience. I'm reading some materials but advice is welcome.

2nd question - broadly speaking, is this even worthwhile when I have only 1 week of data? Am I stuck until I collect more?

related: Least stupid way to forecast a short multivariate time series

Hoping for a bit more explanation or detail, as a time-series beginner. Also that thread doesn't seem to have an accepted answer.


Your Answer

By clicking “Post Your Answer”, you agree to our terms of service, privacy policy and cookie policy

Browse other questions tagged or ask your own question.