Ecosyste.ms: Awesome

An open API service indexing awesome lists of open source software.

Awesome Lists | Featured Topics | Projects

https://github.com/mrc-ide/covid19-forecasts-orderly


https://github.com/mrc-ide/covid19-forecasts-orderly

Last synced: about 2 months ago
JSON representation

Awesome Lists containing this project

README 

        
# orderly



This is an [`orderly`](https://github.com/vimc/orderly) project.  The directories are:



* `src`: create new reports here

* `archive`: versioned results of running your report

* `data`: copies of data used in the reports



Each step of the analysis is an orderly task and corresponds to a

directory on the src directory.



# Description of tasks



1. prepare_ecdc_data This task prepares the ECDC data in a format

   required by the modeling teams. Download the latest data from ECDC

   on Sunday evening and save it as a csv in the folder

   prepare_ecdc_data. Update the resources section in orderly.yml to

   reflect the name of the latest file. Then 

   

   ```

   a <- orderly::orderly_run("prepare_ecdc_data")

   orderly::orderly_commit(a)

   ```

   

   This will generate two artefacts (outputs of a task are called

   artefacts in orderly). They are exactly the same files but one has

   the date in the file name - this file will be shared with the

   modeling groups, and the second file is called

   latest_model_input.rds. This is so that the downstream tasks can

   read the latest file without the need to manually update

   orderly.yml.



2. process_individual_models By Monday afternoon, all modelling teams

   will produce outputs for the report in a prescribed format. These

   could be shared via DropBox for example i.e. the generation of

   model outputs is not part of the orderly workflow. Once these

   outputs are received, save them in a directory on your machine, and

   update the location in orderly_envir.yml. For example, this is the

   file on my machine:

   

   ```

   COVID19_INPUT_PATH: "/Users/sbhatia/OneDrive - Imperial College London/covid19-short-term-forecasts/model_outputs/"

   ```

   Note that the trailing slash in the directory name must be

   included. You can run the task as: 

   

   ```

   a <- orderly::orderly_run("process_individual_models", parameter =

   list(week_ending = "2020-05-31"))

   orderly::orderly_commit(a)

   ```



   produce_performace_metrics Computes various performance metrics 

   for each model.

   

   ```

   a <- orderly::orderly_run("produce_performace_metrics")

   orderly::orderly_commit(a)

   ```



   compute_model_weights Computes model weights using the metrics

   produced in the previous task.

   

3. produce_ensemble_outputs This task creates an ensemble model. 



   

   ```

   a <- orderly::orderly_run("produce_ensemble_outputs", parameters = list(week_ending = "2020-06-14"))

   orderly::orderly_commit(a)

   ```



4. format_model_outputs Pretty formatting of model outputs for

   plugging into the final report.

   

   

   ```

   a <- orderly::orderly_run("format_model_outputs", use_draft = TRUE)

   orderly::orderly_commit(a)

   ```

5. produce_visualisations As the name suggests, makes all the graphs

   needed for the report.



   ```

   a <- orderly::orderly_run("produce_visualisations")

   orderly::orderly_commit(a)

   ```

   

6. produce_exec_summary Generates the summary that goes into the

   report. Run as all other tasks. It is the executive summary which is

   the summary of the new report. produce_summary task is redundant

   and can be ignored for now.



7. produce_full_report Generate the full report.