diff --git a/.github/workflows/data-checks.yml b/.github/workflows/data-checks.yml index d7f07690..f7a2e0ed 100644 --- a/.github/workflows/data-checks.yml +++ b/.github/workflows/data-checks.yml @@ -8,6 +8,9 @@ on: jobs: data-pipeline: runs-on: ubuntu-latest + defaults: + run: + working-directory: data/data-pipeline strategy: matrix: # checks all of the versions allowed in pyproject.toml @@ -20,13 +23,18 @@ jobs: uses: actions/setup-python@v2 with: python-version: ${{ matrix.python-version }} - # installs poetry - - uses: Gr1N/setup-poetry@v4 + - name: Load cached Poetry installation + id: cached-poetry-dependencies + uses: actions/cache@v2 + with: + path: ~/.cache/pypoetry/virtualenvs + key: env-${{ runner.os }}-${{ hashFiles('**/poetry.lock') }}-${{ hashFiles('.github/workflows/data-checks.yml') }} + - name: Install poetry + uses: snok/install-poetry@v1 + - name: Print Poetry settings + run: poetry show -v - name: Install dependencies - run: | - cd data/data-pipeline - poetry install + run: poetry install + if: steps.cached-poetry-dependencies.outputs.cache-hit != 'true' - name: Run tox - run: | - cd data/data-pipeline - poetry run tox + run: poetry run tox diff --git a/data/data-pipeline/data_pipeline/content/config/csv.yml b/data/data-pipeline/data_pipeline/content/config/csv.yml index 85532f7a..2d9764b0 100644 --- a/data/data-pipeline/data_pipeline/content/config/csv.yml +++ b/data/data-pipeline/data_pipeline/content/config/csv.yml @@ -18,7 +18,7 @@ fields: label: Total threshold criteria exceeded format: int64 - score_name: Definition M (communities) - label: Definition M (communities) + label: Identified as disadvantaged format: bool - score_name: Total population label: Total population diff --git a/data/data-pipeline/data_pipeline/content/config/excel.yml b/data/data-pipeline/data_pipeline/content/config/excel.yml index 913de7c6..e9f77ba1 100644 --- a/data/data-pipeline/data_pipeline/content/config/excel.yml +++ b/data/data-pipeline/data_pipeline/content/config/excel.yml @@ -23,7 +23,7 @@ sheets: label: Total threshold criteria exceeded format: int64 - score_name: Definition M (communities) - label: Definition M (communities) + label: Identified as disadvantaged format: bool - score_name: Total population label: Total population diff --git a/data/data-pipeline/data_pipeline/etl/score/tests/snapshots/downloadable_data_expected.pkl b/data/data-pipeline/data_pipeline/etl/score/tests/snapshots/downloadable_data_expected.pkl index ae1cfaf0..edaaa2ae 100644 Binary files a/data/data-pipeline/data_pipeline/etl/score/tests/snapshots/downloadable_data_expected.pkl and b/data/data-pipeline/data_pipeline/etl/score/tests/snapshots/downloadable_data_expected.pkl differ