Make aggregation field types consistent (#66)

* AH-1801 Change key dtypes to match
* AH-1801 update workflow file - Trying fixes from conda-incubator/setup-miniconda#274
* AH-1801 fix pylint errors

.github/workflows/hazimp-tests.yml

@@ -13,12 +13,15 @@ jobs:
   Hazimp:
     name: Test HazImp
     runs-on: ubuntu-latest
-    strategy: 
+    strategy:
       matrix:
         python-version: [3.7, 3.8, 3.9]
+    defaults:
+      run:
+        shell: bash -l {0}
 
     steps:
-    - uses: actions/checkout@v2
+    - uses: actions/checkout@v3
     - name: Cache conda
       uses: actions/cache@v2
       env:
@@ -27,13 +30,14 @@ jobs:
         path: ~/conda_pkgs_dir
         key:
           ${{ runner.os }}-conda-${{ env.CACHE_NUMBER }}-${{hashFiles('hazimp.yml') }}
-          
+
     - name: Set up environment
-      uses: conda-incubator/setup-miniconda@v2.0.0
+      uses: conda-incubator/setup-miniconda@v2
       with:
-        python-version: ${{ matrix.python-version }}
+        miniforge-variant: Mambaforge
         channels: conda-forge,defaults
-        mamba-version: "*"
+        channel-priority: true
+        python-version: ${{ matrix.python-version }}
         activate-environment: hazimp
         environment-file: hazimp.yml
         auto-activate-base: false

hazimp/aggregate.py

@@ -9,7 +9,7 @@
 import geopandas
 import pandas as pd
 import numpy as np
-
+from hazimp.misc import check_data_type
 LOGGER = logging.getLogger(__name__)
 
 # List of possible drivers for output:
@@ -91,6 +91,19 @@ def choropleth(dframe, boundaries, impactcode, bcode, filename,
 
     left, right = impactcode, bcode
 
+    shapes = geopandas.read_file(boundaries)
+    try:
+        dtype = check_data_type(shapes[right])
+    except KeyError:
+        LOGGER.exception(f"Aggregation boundaries have no attribute '{right}'")
+        sys.exit(1)
+
+    try:
+        dframe[left] = dframe[left].astype(dtype)
+    except ValueError:
+        LOGGER.exception(f"Cannot convert {left} to {dtype}")
+        sys.exit(1)
+
     aggregate = dframe.groupby(left).agg(fields).reset_index()
     aggregate.columns = [
         '_'.join(columns).rstrip('_') for columns in aggregate.columns.values
@@ -111,18 +124,7 @@ def choropleth(dframe, boundaries, impactcode, bcode, filename,
     else:
         aggregate.set_index(left, inplace=True)
 
-    shapes = geopandas.read_file(boundaries)
-
-    try:
-        shapes['key'] = shapes[right].astype(np.int64)
-    except KeyError:
-        LOGGER.error(f"{boundaries} does not contain an attribute {right}")
-        sys.exit(1)
-    except OverflowError:
-        LOGGER.error(f"Unable to convert {right} values to ints")
-        sys.exit(1)
-
-    result = shapes.merge(aggregate, left_on='key', right_index=True)
+    result = shapes.merge(aggregate, left_on=right, right_index=True)
 
     fileext = os.path.splitext(filename)[1].replace('.', '')
     try:

hazimp/misc.py

@@ -454,3 +454,18 @@ def upload_to_s3_if_applicable(local_path, bucket_name, bucket_key,
         if not ignore_exception:
             LOGGER.exception("S3 write error: {0}".format(local_path))
             raise e
+
+
+def check_data_type(data):
+    """
+    Function to check the data type of a given attribute
+
+    :param data: Sample of the data
+    :type data: `pd.Series` or `pd.DataFrame`
+    """
+    try:
+        dtype = data[0].dtype
+    except AttributeError:
+        dtype = type(data[0])
+
+    return dtype