🐛 Remove Google Sheet from sources of fast-tracked datasets (#3618)

* 🐛 Remove Google Sheet from sources of fast-tracked datasets

apps/wizard/app_pages/fasttrack/load.py

@@ -294,7 +294,7 @@ def _parse_sources(sources_meta_df: pd.DataFrame) -> Optional[Source]:
     source = sources[0]
 
     if pd.isnull(source.get("date_accessed")):
-        source.pop("date_accessed")
+        source.pop("date_accessed", None)
 
     if pd.isnull(source.get("publication_year")):
         source.pop("publication_year")

apps/wizard/app_pages/fasttrack/process.py

@@ -151,7 +151,7 @@ def _convert_percentages(data: pd.DataFrame, variables_meta_dict: Dict[str, Vari
     """Convert percentages to numbers."""
     for col in data.columns:
         if getattr(variables_meta_dict.get(col, {}), "unit", "") == "%":
-            data[col] = data[col].str.replace("%", "").astype(float)
+            data[col] = data[col].astype(str).str.replace("%", "").astype(float)
     return data
 
 

etl/steps/data/grapher/fasttrack/2022-11-01/lighting_efficiency_uk.meta.yml

@@ -1,40 +1,18 @@
 dataset:
-  namespace: fasttrack
-  version: '2022-11-01'
-  short_name: lighting_efficiency_uk
   title: Lighting effiency and shares in the UK
   description: ''
-  sources:
-  - name: 'Fouquet & Pearson (2006). Seven centuries of energy services: The price
-      and use of light in the United Kingdom (1300-2000).'
-    published_by: 'Fouquet, R., & Pearson, P. J. (2006). Seven centuries of energy
-      services: The price and use of light in the United Kingdom (1300-2000). The
-      energy journal, 27(1).'
-    url: https://www.jstor.org/stable/23296980
+  licenses:
+    - {}
 tables:
   lighting_efficiency_uk:
     variables:
       share_of_lighting_uk:
         title: share_of_lighting_uk
-        short_unit: '%'
         unit: '%'
+        short_unit: '%'
         description: The share of lighting in the UK that was provided by each source.
-        sources:
-        - name: 'Fouquet & Pearson (2006). Seven centuries of energy services: The
-            price and use of light in the United Kingdom (1300-2000).'
-          published_by: 'Fouquet, R., & Pearson, P. J. (2006). Seven centuries of
-            energy services: The price and use of light in the United Kingdom (1300-2000).
-            The energy journal, 27(1).'
-          url: https://www.jstor.org/stable/23296980
       efficiency_lighting_uk:
         title: efficiency_lighting_uk
         unit: lumen-hours per kWh
-        description: The efficiency of lighting measures the output of light per unit
-          of energy. It's measured in lumen-hours per kilowatt-hour (kWh).
-        sources:
-        - name: 'Fouquet & Pearson (2006). Seven centuries of energy services: The
-            price and use of light in the United Kingdom (1300-2000).'
-          published_by: 'Fouquet, R., & Pearson, P. J. (2006). Seven centuries of
-            energy services: The price and use of light in the United Kingdom (1300-2000).
-            The energy journal, 27(1).'
-          url: https://www.jstor.org/stable/23296980
+        description: |-
+          The efficiency of lighting measures the output of light per unit of energy. It's measured in lumen-hours per kilowatt-hour (kWh).

etl/steps/data/grapher/fasttrack/2022-11-01/lighting_efficiency_uk.py

@@ -1,21 +1,40 @@
 import pandas as pd
-from owid import catalog
 
-from etl.helpers import PathFinder
+from etl.helpers import PathFinder, create_dataset, get_metadata_path
 from etl.snapshot import Snapshot
 
-N = PathFinder(__file__)
+paths = PathFinder(__file__)
 
 
 def run(dest_dir: str) -> None:
     # load snapshot
-    data = pd.read_csv(Snapshot("fasttrack/2022-11-01/lighting_efficiency_uk.csv").path)
+    snap = Snapshot("fasttrack/2022-11-01/lighting_efficiency_uk.csv")
 
-    # create empty dataframe and table
-    ds = catalog.Dataset.create_empty(dest_dir)
-    tb = catalog.Table(data, short_name=N.short_name)
+    # load data
+    tb = snap.read_csv()
+
+    # add dimensions with dim_ prefix
+    dims = [c for c in tb.columns if c.startswith("dim_")]
+    dims_without_prefix = [c[4:] for c in dims]
+
+    if dims:
+        tb = tb.rename(columns={d: dw for d, dw in zip(dims, dims_without_prefix)})
+
+    if uses_dates(tb["year"]):
+        tb = tb.rename(columns={"year": "date"}).format(["country", "date"] + dims_without_prefix)
+    else:
+        tb = tb.format(["country", "year"] + dims_without_prefix)
 
     # add table, update metadata from *.meta.yml and save
-    ds.add(tb)
-    ds.update_metadata(N.metadata_path)
+    ds = create_dataset(dest_dir, tables=[tb], default_metadata=snap.metadata)
+
+    # override metadata if necessary
+    meta_path = get_metadata_path(dest_dir).with_suffix(".override.yml")
+    if meta_path.exists():
+        ds.update_metadata(meta_path)
+
     ds.save()
+
+
+def uses_dates(s: pd.Series) -> bool:
+    return pd.to_datetime(s, errors="coerce", format="%Y-%m-%d").notnull().all()