This is a guide on what to do after you have created a new task repository from the template. More in depth information about how to create a new task can be found in the OpenProblems Documentation.
- Update the
_viash.yaml
file with the correct task information. - Update the
src/api/task_info.yaml
file with the information you have provied in the task issue.
THe OpenProblems team has provided some test resources that can be used to test the task. These resources are stored in the resources
folder. The scripts/download_resources.sh
script can be used to download these resources.
If these resources are not sufficient, you can add more resources to the resources
folder. The scripts/download_resources.sh
script can be updated to download these resources.
#!/bin/bash
echo "This script is not supposed to be run directly." echo "Please run the script step-by-step." exit 1
scripts/download_resources.sh
method_id="my_metric" method_lang="python" # change this to "r" if need be
common/create_component/create_component --
--language "$method_lang"
--name "$method_id"
viash test src/task/methods/$method_id/config.vsh.yaml
You can reduce the memory and cpu allotted to jobs in _viash.yaml by modifying .platforms[.type == "nextflow"].config.labels
viash run src/task/methods/$method_id/config.vsh.yaml --
---setup cachedbuild ---verbose
viash run src/task/methods/$method_id/config.vsh.yaml --
--de_train "resources/neurips-2023-kaggle/de_train.parquet"
--id_map "resources/neurips-2023-kaggle/id_map.csv"
--output "output/prediction.parquet"
viash run src/task/methods/$method_id/config.vsh.yaml --
--de_train_h5ad "resources/neurips-2023-kaggle/2023-09-12_de_by_cell_type_train.h5ad"
--id_map "resources/neurips-2023-kaggle/id_map.csv"
--output "output/prediction.parquet"
viash run src/task/metrics/mean_rowwise_error/config.vsh.yaml --
--de_test "resources/neurips-2023-kaggle/de_test.parquet"
--prediction "output/prediction.parquet"
--output "output/score.h5ad"
python -c 'import anndata; print(anndata.read_h5ad("output/score.h5ad").uns)'