witch-team / witchtools Goto Github PK

View Code? Open in Web Editor NEW

5.0 9.0 2.0 7.71 MB

Data management tools for the WITCH modeling framework

Home Page: https://witchtools.witchmodel.org

License: Other

R 40.09% Assembly 56.64% Pawn 1.39% C++ 1.87%

r package integrated-assessment-model

witchtools's Introduction

witchtools

witchtools provides a toolkit to manage data in integrated assessment models. witchtools is part of the tool ecosystem of the WITCH model. The package makes it possible to easily and transparently reproduce the data processing between the raw data, the data used in the model and the results uploaded in the IAMC database. While the package has been developed for the WITCH model, it is already used to simply process data for other uses or models.

Features

Conversion of a yearly time series into a time-period (including possible decision tree structure) based time series.
Conversion of data from one regional mapping to another one.
Default time-period mappings, regional mappings and aggregation weights used by the WITCH model
Batch conversion of the make_data files from the input directory of the WITCH model
Easy access to the witch-team/witch-data repository

Installation

if (!requireNamespace("remotes"))
  install.packages("remotes")

remotes::install_github("witch-team/witchtools")

Usage

library(witchtools)

# Convert from 17 WITCH regions to the 5 SSP regions 
gdp_r5 <- convert_region(gdp_witch17, to_reg = 'r5')

# Convert from yearly to the WITCH time perdios
dd_t30 <- convert_region(dd_year, time_mapping = 't30')

# for WITCH user: Generate input data of the WITCH model
witch_translate_data(region = 'witch17', timescale = 't30')

witchtools's People

Contributors

Stargazers

Watchers

Forkers

greentea6 camillamoioli

witchtools's Issues

weights for 'gdp' is defined twice

Should we use the SSP source or the IMF source?

two questions

Thank you Johannes for the youtube link, it was very helpful. And, apologies in advance for basic questions.

Is it possible to do a country specific analysis?
is it possible to have carbon tax assigned to a specific sector, for example, coal? - also is it possible to do it in only one country [example USA]and keep the rest the same. I am assuming this will need code changes in the core policy file. any guidance on them would be helpful.

dvc pull needs quote for filenames

allow changing region mappings

Right now the region mappings seem hard coded and loaded in the packaged in an rda file.
We should think about a way to allow user specific mappings.
Both for other users, and also for us since now the .inc files used in witch and these hard coded sets are pretty incompatible, or we add a function to add or convert user defined mappings?

Debug mode & separation from witch-data

A few things from my experience in working with the witchtools package:

I think it would be really important to have a way in which new users can easily edit their files locally of the whole witch-* framework without having always to use github, ie using witch-data and witchtools in the local folder
Given it is an independent package would be good to make it work also based on a completely new witch-mode/input and witch-data folder. However, as of now it doesn't work: If I create all the files (ie run translate_witch_data), and then switch to witch-data method on an empty folder, and just rerun the script it gives a few errors: (Yet in my opinion, this should work without errors).
-1) ssp data not found (OK I assume because of the weights).
-2) "database connection not found" not sure which one (globiom? primap for emission weights?).

I think the weights we should or embed them in the package or make them by default use piggyback so it will work with a completely new data folder for other models etc.

Anyhow we should discuss in a witch strategic meeting!

region/time mapping outside the package

So ther eis reallt no way to have the region/time inc files not in the packages' too bad.

Correct definition of EU

To be decided, but probably the set EU shoujdl contain only the current valid members of the European Union.
that is, 27 countries at present.
It seems in the moment other regions are in there like UK or so, since othereurope in witch20 is flagged as being also EU.
Also, should be based on population majority to wether it belongs to eht EU set or not imho.

Consistent time mappings and allow adding new ones

Now regional mappings are OK to added. SO would make sense to also allow custom time mappings.
In the pre-Package version there was the function
load_time_mapping = function(file){ tab = fread(file,colClasses="character") return(tab) }

However now I see some issues

There is the new column "refyear" which is NOT in the raw csv files neither in WITCH nor witchtools
the csv files are on yearly basis not tstep, so the function woudl be more complex now, but let's discuss how to have consisteny here too.

Convert_region additional check

Add a check to make sure all the members of the list passed to the regions field of convert_region are of type data.table and send an error message if they are not

update time and region mappings

I would suggest to add the DICE t58 mapping, ed57 (and remove enerdata56 which are actually 57 regions and it is nowhere used anymore) and add maxiso3 (from RICE50x inc files).