Merge pull request #164 from statisticsnorway/valuta_timeperiod_formats

Valuta timeperiod formats

Author: aecorn

demos/valuta_check.py

@@ -5,7 +5,7 @@
 #       extension: .py
 #       format_name: percent
 #       format_version: '1.3'
-#       jupytext_version: 1.15.2
+#       jupytext_version: 1.17.2
 #   kernelspec:
 #     display_name: ssb-fagfunksjoner
 #     language: python
@@ -19,12 +19,11 @@
 print(fagfunksjoner.__version__)
 
 # %%
-from IPython.display import display
 
 from fagfunksjoner.api.valuta import download_exchange_rates
 
 
 exchange_rates = download_exchange_rates()
-display(exchange_rates.df)
+exchange_rates.df  # noqa: B018
 
 # %%

pyproject.toml

@@ -1,6 +1,6 @@
 [tool.poetry]
 name = "ssb-fagfunksjoner"
-version = "1.1.1"
+version = "1.1.2"
 description = "Fellesfunksjoner for ssb i Python"
 authors = ["SSB-pythonistas <ssb-pythonistas@ssb.no>"]
 license = "MIT"

src/fagfunksjoner/api/valuta.py

@@ -1,9 +1,10 @@
+import datetime
 from dataclasses import dataclass
-from datetime import date
 from typing import Any
 
 import pandas as pd
 import requests
+from dateutil import parser
 
 
 @dataclass
@@ -287,10 +288,6 @@ def create_dataframe(data_obj: Data, structure_obj: Structure) -> pd.DataFrame:
     Returns:
         pd.DataFrame: A pandas DataFrame created from the data and structure objects.
     """
-    dimension_keys = [dim.id for dim in structure_obj.dimensions.get("series", [])]
-    observation_keys = [
-        dim.id for dim in structure_obj.dimensions.get("observation", [])
-    ]
     records = []
 
     for dataset_obj in data_obj.dataSets:
@@ -301,25 +298,7 @@ def create_dataframe(data_obj: Data, structure_obj: Structure) -> pd.DataFrame:
                         series_key, series_val, obs_key, obs_value, structure_obj
                     )
                 )
-
-    return pd.DataFrame(
-        records,
-        columns=dimension_keys
-        + [dim.id + "_id" for dim in structure_obj.dimensions.get("series", [])]
-        + observation_keys
-        + [dim.id + "_id" for dim in structure_obj.dimensions.get("observation", [])]
-        + ["Observation"]
-        + [
-            attr.id
-            for attr_list in structure_obj.attributes.values()
-            for attr in attr_list
-        ]
-        + [
-            attr.id + "_id"
-            for attr_list in structure_obj.attributes.values()
-            for attr in attr_list
-        ],
-    )
+    return pd.DataFrame(records)
 
 
 def make_single_dataframe_record(
@@ -328,7 +307,7 @@ def make_single_dataframe_record(
     obs_key: str,
     obs_value: list[str],
     structure_obj: Structure,
-) -> dict[str, str | float]:
+) -> dict[str, str | float | datetime.datetime]:
     """Create a single record for a pandas DataFrame from a series and observation.
 
     This function generates a dictionary representing a single row in a pandas
@@ -345,7 +324,16 @@ def make_single_dataframe_record(
     Returns:
         dict[str, str | float]: A dictionary representing a single record in the DataFrame.
     """
-    record: dict[str, str | float] = {
+    observation_fields: dict[str, str | datetime.datetime] = {}
+    for dim in structure_obj.dimensions.get("observation", []):
+        observation: dict[str, str] = next(
+            (val for i, val in enumerate(dim.values) if i == int(obs_key))
+        )
+        observation_fields |= {
+            f"{dim.id}_{field}": parser.parse(val) for field, val in observation.items()
+        }
+
+    record: dict[str, str | float | datetime.datetime] = {
         **{
             dim.id: dim.values[int(series_key.split(":")[dim.keyPosition])]["name"]
             for dim in structure_obj.dimensions.get("series", [])
@@ -355,17 +343,7 @@ def make_single_dataframe_record(
             + "_id": dim.values[int(series_key.split(":")[dim.keyPosition])]["id"]
             for dim in structure_obj.dimensions.get("series", [])
         },
-        **{
-            dim.id: next(
-                (val["name"] for val in dim.values if val["id"] == obs_key),
-                obs_key,
-            )
-            for dim in structure_obj.dimensions.get("observation", [])
-        },
-        **{
-            dim.id + "_id": obs_key
-            for dim in structure_obj.dimensions.get("observation", [])
-        },
+        **observation_fields,
         "Observation": obs_value[0],
     }
     for _attr_key, attr_list in structure_obj.attributes.items():
@@ -423,6 +401,7 @@ def parse_response(json_data: dict[str, Any]) -> ValutaData:
 
     # Create DataFrame
     df = create_dataframe(data_obj, structure_obj)
+
     valuta_data = ValutaData(meta=valuta_meta, data=data_obj, df=df)
 
     return valuta_data
@@ -458,7 +437,7 @@ def download_exchange_rates(
         ValutaData: The data retrieved from the API, parsed into a ValutaData object.
     """
     if date_to is None:
-        date_to = date.today().strftime("%Y-%m-%d")
+        date_to = datetime.date.today().strftime("%Y-%m-%d")
 
     url = URL_NORGES_BANK.format(
         frequency=frequency,
@@ -472,5 +451,4 @@ def download_exchange_rates(
     response = requests.get(url)
     response.raise_for_status()
     json_data = response.json()
-
     return parse_response(json_data)

tests/api/test_valuta.py

@@ -209,20 +209,22 @@ def test_parse_response():
             "BASE_CUR",
             "QUOTE_CUR",
             "TENOR",
-            "TIME_PERIOD",
             "FREQ_id",
             "BASE_CUR_id",
             "QUOTE_CUR_id",
             "TENOR_id",
+            "TIME_PERIOD_start",
+            "TIME_PERIOD_end",
             "TIME_PERIOD_id",
+            "TIME_PERIOD_name",
             "Observation",
             "DECIMALS",
-            "CALCULATED",
-            "UNIT_MULT",
-            "COLLECTION",
             "DECIMALS_id",
+            "CALCULATED",
             "CALCULATED_id",
+            "UNIT_MULT",
             "UNIT_MULT_id",
+            "COLLECTION",
             "COLLECTION_id",
         }
     )