Comments (5)
I added model-id-map
files in each of the three model type directories.
A sample is here.
model-id,model-dir
CU-BMA,CUBMA
CU-EAKFC,CUEAKFC
Delphi-BasisRegression,Delphi_BasisRegression_PackageDefaults
Delphi-DeltaDensity1,Delphi_DeltaDensity_PackageDefaults
...
The first column is model id (team_abbr-model_abbr
) and next is the folder (adjacent to the csv. I can change it to the complete path relative to repo root if needed). These mappings are not per season and doesn't have any sense about which models are complete in any season.
from cdc-flusight-ensemble.
Update:
I added code that skips model mapping for incomplete models. This means that the new file (which is complete-modelids.csv
now) has three columns:
season,model-id,model-dir
2010/2011,Delphi-BasisRegression,Delphi_BasisRegression_PackageDefaults
2010/2011,Delphi-DeltaDensity1,Delphi_DeltaDensity_PackageDefaults
2010/2011,Delphi-EmpiricalBayes1,Delphi_EmpiricalBayes_Cond4
2010/2011,Delphi-EmpiricalBayes2,Delphi_EmpiricalBayes_PackageDefaults
...
If a model hasn't provided submission files from week 40 to week 20 (inclusive) for each season, then that doesn't get mentioned in that season. This is a harsh test and some models get skipped completely (CU*, RL_KCDE). Maybe there is some off by one thing? All of the models are not submitting exactly 33 (or 34 for season with 53 weeks) files per season?
from cdc-flusight-ensemble.
There are some missing files, so this is not too harsh of a test.
from cdc-flusight-ensemble.
Would it be possible, for now, to include both versions of this file? The model-id-map
and the complete-modelids
? The score calculations do not ignore incomplete models, so we need the map for all models. But seeing which are complete is really useful too. Maybe we could use model-id-map
and add a single column with an 0/1 indicator for "complete". where complete means it has all necessary model files for all years, without the year-by-year breakdown.
from cdc-flusight-ensemble.
Added the automation script for model-id-map.csv
file. This file now also has column specifying whether the model is complete (in all the seasons).
from cdc-flusight-ensemble.
Related Issues (20)
- Automerge and push from travis HOT 1
- point estimates scores use calendar week and not model week HOT 1
- calculate new weights HOT 1
- remove DBM model
- re-score all models
- re-generate targets
- check that re-build of CV-ensembles using updated models works
- remove UTaustin model
- create final weight files for season
- create final CV ensemble files for 2018/2019 season
- re-order protea kudu model columns
- are historical ILI rates (known at time T) available? HOT 1
- Confirm that target-multivals are created based on data as of the end of season HOT 1
- Incomplete forecast file HOT 5
- error with at least 1 UA forecast file HOT 3
- Error in EW49 2018 CU_EAKFC_SIRS data format HOT 3
- Error in Protea_Kudu Forecast Formatting HOT 3
- Protea Kudu Formatting Edits
- Protea_Kudu forecasts have unexpected indentation in the βvalueβ column HOT 3
- Update flusight-deploy to not depend on the non-updated FluSight repository
Recommend Projects
-
React
A declarative, efficient, and flexible JavaScript library for building user interfaces.
-
Vue.js
π 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. πππ
-
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 β€οΈ Open Source for everyone.
-
Alibaba
Alibaba Open Source for everyone
-
D3
Data-Driven Documents codes.
-
Tencent
China tencent open source team.
from cdc-flusight-ensemble.