def load_schema(self, indicator):
df = indicator.data
schema = describe_schema(df)
non_disaggregation_columns = indicator.options.get_non_disaggregation_columns(
)
non_disaggregation_special_columns = [
indicator.options.get_series_column(),
indicator.options.get_unit_column(),
]
for column in indicator.data.columns:
if column in non_disaggregation_columns and column not in non_disaggregation_special_columns:
continue
unique = df[column].unique()
unique = unique[~pd.isnull(unique)]
unique = unique.tolist() + ['']
schema.get_field(column).constraints['enum'] = unique
return schema
def test_schema_to_copy():
source = describe_schema("data/table.csv")
target = source.to_copy()
assert source is not target
assert source == target
def test_describe_schema():
schema = describe_schema("data/leading-zeros.csv")
assert schema == {"fields": [{"name": "value", "type": "integer"}]}