search

Weiming-Hu / AnalogsEnsemble

The C++ and R packages for parallel ensemble forecasts using Analog Ensemble
https://weiming-hu.github.io/AnalogsEnsemble/
MIT License
18 stars 5 forks source link

Work with timeseries of observations and forecast #108

Closed thanasisn closed 4 years ago

thanasisn commented 4 years ago

I am using RAnEn to correct forecast using observations.

I have a minimum example of my implementation that should run easily: https://github.com/natsis/RAnEn_application I think I have structured the data correctly but still, I am missing something.

The scripts runs, but every time, I can get only one prediction per day and not the hourly values I expected.

thanasisn commented 4 years ago

I had an follow up answer, as I fist ask through email

I took a look at your script. Your observations turn out to be mostly NA. This is probably why your analogs are mostly NA. Do you know why your observations are mostly NA?

> length(observations$Data)
[1] 34632
> length(which(is.na(observations$Data)))
[1] 33270
thanasisn commented 4 years ago

Thanks to Weiming-Hu I solved the problem! The reshaping of the data was not correct!

Weiming-Hu commented 4 years ago

Thank you for taking care of this.