diff --git a/.github/workflows/parse_catalog.yaml b/.github/workflows/parse_catalog.yaml index d884483b..b8fb4549 100644 --- a/.github/workflows/parse_catalog.yaml +++ b/.github/workflows/parse_catalog.yaml @@ -25,7 +25,7 @@ jobs: python-version: "3.9" - name: Install dependencies # TODO: base on an actual release - run: python -m pip install gcsfs git+https://github.com/leap-stc/leap-data-management-utils.git@factor-out-cmip + run: python -m pip install gcsfs "leap-data-management-utils[bigquery]==0.0.14" - name: "Parse catalog" shell: bash run: | diff --git a/scripts/dump_bigquery_to_csv.py b/scripts/dump_bigquery_to_csv.py index 7b559158..44157e88 100644 --- a/scripts/dump_bigquery_to_csv.py +++ b/scripts/dump_bigquery_to_csv.py @@ -1,4 +1,4 @@ -from leap_data_management_utils import CMIPBQInterface +from leap_data_management_utils.bq_interfaces import CMIPBQInterface from leap_data_management_utils.cmip_catalog import bq_df_to_intake_esm import os import gcsfs @@ -32,6 +32,10 @@ ) if len(bq_df) > 0: + print(df_all.head().columns) + print(df_all.head().instance_id.tolist()) + print(df_all.head().store.tolist()) + intake_esm_df = bq_df_to_intake_esm(bq_df) intake_esm_df.to_csv(filename, index=False) if fs.exists(path):