Comments (1)
Not actually an issue - mean fields are added to the residuals in lines 100-102 https://github.com/JGCRI/fldgen/blob/master/R/generateTPfullgrids.R#L100-L102
and then that whole thing is what gets converted to native units in https://github.com/JGCRI/fldgen/blob/master/R/generateTPfullgrids.R#L106-L123
In trainTP
, values get converted (https://github.com/JGCRI/fldgen/blob/master/R/trainTP.R#L149-L166) before the mean field is calculated (https://github.com/JGCRI/fldgen/blob/master/R/trainTP.R#L188-L192).
Precip as an example:
We convert full precip to full logP (https://github.com/JGCRI/fldgen/blob/master/R/trainTP.R#L160-L166). We take the mean field and residuals from that data (https://github.com/JGCRI/fldgen/blob/master/R/trainTP.R#L192), so the mean field stored in the trained emulator is in units of logP and the residuals generated are in units of logP. In other words
That means the reconstruction function called on lines 89 and 90 of generateTPfullgrids
is operating on logP and resulting in logP mean fields.
So when we add residuals and mean fields together on lines 100-102, we have logP full fields. When we go back to native units on line 106-123, the operation being done is then
from fldgen.
Related Issues (20)
- clean up comments and move internal functions in normalizeresiduals.R to be external HOT 1
- fix the memory bloat of a saved emulator HOT 6
- downsized data sets to speed up testing HOT 1
- Create and archive trained emulators for popular ESMs HOT 2
- Instructions in tutorial2 are incorrect
- Add option to splice a historical scenario onto each future scenario
- Misleading documentation in generate.TP.fullgrids
- Quantile transform sometimes generates NA values
- globalop field in griddata is incorrect when grid cells are missing
- Check with full GCM data - impacts of training on full data or just land data
- Add github actions to the repo HOT 2
- Add package down to fldgen HOT 1
- 'reading and writing models works' test fails HOT 6
- remove dependency on gcammaptools HOT 1
- Problem with testing fldgen on developmental version of R HOT 8
- update stickers on the readme HOT 1
- variability in Tgav and using Hector to drive trained ESM emulators HOT 1
- Add tutorials to actually go through emulator validation HOT 2
- revisit warnings - variable labels - units HOT 1
Recommend Projects
-
React
A declarative, efficient, and flexible JavaScript library for building user interfaces.
-
Vue.js
🖖 Vue.js is a progressive, incrementally-adoptable JavaScript framework for building UI on the web.
-
Typescript
TypeScript is a superset of JavaScript that compiles to clean JavaScript output.
-
TensorFlow
An Open Source Machine Learning Framework for Everyone
-
Django
The Web framework for perfectionists with deadlines.
-
Laravel
A PHP framework for web artisans
-
D3
Bring data to life with SVG, Canvas and HTML. 📊📈🎉
-
Recommend Topics
-
javascript
JavaScript (JS) is a lightweight interpreted programming language with first-class functions.
-
web
Some thing interesting about web. New door for the world.
-
server
A server is a program made to process requests and deliver data to clients.
-
Machine learning
Machine learning is a way of modeling and interpreting data that allows a piece of software to respond intelligently.
-
Visualization
Some thing interesting about visualization, use data art
-
Game
Some thing interesting about game, make everyone happy.
Recommend Org
-
Facebook
We are working to build community through open source technology. NB: members must have two-factor auth.
-
Microsoft
Open source projects and samples from Microsoft.
-
Google
Google ❤️ Open Source for everyone.
-
Alibaba
Alibaba Open Source for everyone
-
D3
Data-Driven Documents codes.
-
Tencent
China tencent open source team.
from fldgen.