Select agencies (#23)

* move function to utils

* adjust input agencies

* use acronyms instead of slugs

for selection menu

* remove acronyms

keeps improved get metadata values function; values in menu now scrollable via css

app.py

@@ -271,6 +271,8 @@ with ui.accordion(open=False):
             This dashboard allows users to explore how different lookback window dates would affect the set of rules available for congressional review. 
             
             Rule data are retrieved daily from the [Federal Register API](https://www.federalregister.gov/developers/documentation/api/v1), which publishes new editions of the Federal Register each business day. 
+
+            A list of common agency acronyms is available from the [U.S. Government Manual](https://www.govinfo.gov/content/pkg/GOVMAN-2022-12-31/pdf/GOVMAN-2022-12-31-Commonly-Used-Acronyms-105.pdf).
             """
             )
 
@@ -284,7 +286,7 @@ ui.markdown(
 
 
 @reactive.calc
-def filtered_df():
+def filtered_df(agency_column: str = "parent_slug"):
     filt_df = DF
     
     # filter dates
@@ -295,7 +297,7 @@ def filtered_df():
     
     # filter agencies
     if (input.menu_agency() is not None) and ("all" not in input.menu_agency()):
-        bool_agency = [True if sum(selected in agency for selected in input.menu_agency()) > 0 else False for agency in filt_df["parent_slug"]]
+        bool_agency = [True if sum(selected in agency for selected in input.menu_agency()) > 0 else False for agency in filt_df[agency_column]]
         filt_df = filt_df.loc[bool_agency]
     
     # return filtered dataframe

modules/__init__.py

@@ -3,6 +3,7 @@ from .grouping import *
 from .plotting import *
 from .search_columns import *
 from .significant import *
+from .utils import *
 
 
 # see: https://docs.python.org/3.11/tutorial/modules.html#packages
@@ -12,4 +13,5 @@ __all__ = [
     "plotting",
     "search_columns", 
     "significant", 
+    "utils", 
     ]

modules/get_rules_in_window.py

@@ -8,9 +8,11 @@ from pandas import DataFrame, to_datetime
 try:
     from search_columns import search_columns, SearchError
     from significant import get_significant_info
+    from utils import get_agency_metadata_values
 except (ModuleNotFoundError, ImportError):
     from .search_columns import search_columns, SearchError
     from .significant import get_significant_info
+    from .utils import get_agency_metadata_values
 
 
 METADATA, _ = AgencyMetadata().get_agency_metadata()
@@ -20,14 +22,16 @@ GET_SIGNIFICANT = True if date.fromisoformat(START_DATE) >= date(2023, 4, 6) els
 
 
 class DataAvailabilityError(Exception):
+    """Raised when data is not available for the requested inputs."""
     pass
 
 
-def get_date_range(start_date: str):
+def get_date_range(start_date: str, end_mmdd: str = "01-03"):
     """Define date range of documents returned by the app.
 
     Args:
         start_date (str): The start date for retrieving the documents.
+        end_mmdd (str, optional): The month and day for the end date in MM-DD format. Defaults to "01-03".
 
     Returns:
         dict: Dictionary containing start date, end date, and transition year.
@@ -36,7 +40,7 @@ def get_date_range(start_date: str):
     end_year = start_year + 1
     date_range = {
         "start": start_date, 
-        "end": f"{end_year}-01-03", 
+        "end": f"{end_year}-{end_mmdd}", 
         "transition_year": end_year, 
         }
     return date_range
@@ -155,13 +159,14 @@ def get_significant_rules(df, start_date):
     return df, last_updated
 
 
-def get_rules_in_window(start_date: str, get_significant: bool = True):
+def get_rules_in_window(start_date: str, get_significant: bool = True, metadata=METADATA):
     date_range = get_date_range(start_date)
     transition_year = date_range.get("transition_year")
     results = get_rules(date_range)
     df = format_documents(results)
     df, _ = filter_corrections(df)
     df = filter_new_admin_rules(df, transition_year)
+    df.loc[:, "acronym"] = get_agency_metadata_values(df, "parent_slug", metadata=METADATA, metadata_value="acronym")
     if get_significant:
         df, last_updated = get_significant_rules(df, start_date)
     else:
@@ -184,3 +189,4 @@ if __name__ == "__main__":
     print(DF.columns)
     print(LAST_UPDATED)
     print(AGENCIES)
+    print(len(METADATA.keys()))

modules/grouping.py

@@ -5,29 +5,7 @@ from datetime import datetime, date, timedelta
 from dateutil.relativedelta import *
 from pandas import DataFrame, Timestamp, to_datetime
 
-
-def _get_agency_metadata_values(
-        df: DataFrame, 
-        agency_column: str, 
-        metadata: dict,         
-        metadata_value: str, 
-    ):
-    """Get a specific value from agency metadata (e.g., get acronym for Department of Homeland Security).
-
-    Args:
-        df (DataFrame): Input data.
-        agency_column (str): Column containing agency identifier.
-        metadata (dict): Agency metadata.
-        metadata_value (str): Value of interest from agency metadata.
-
-    Returns:
-        pd.Series: Pandas Series of new values for adding to DataFrame.
-    """    
-    if metadata_value == "acronym":
-        metadata_value = "short_name"
-    return df.loc[:, agency_column].apply(
-        lambda x: metadata.get(x, {}).get(metadata_value)
-        )
+from .utils import get_agency_metadata_values
 
 
 def _get_first_week_start(dates: list[date], week_start: int | str | "weekday" = MO):
@@ -229,7 +207,7 @@ def groupby_agency(
             }, errors="ignore"
         )
     if metadata is not None:
-        grouped.loc[:, metadata_value] = _get_agency_metadata_values(
+        grouped.loc[:, metadata_value] = get_agency_metadata_values(
         grouped, 
         agency_column="agency", 
         metadata=metadata, 

modules/utils.py

@@ -0,0 +1,34 @@
+from pandas import DataFrame
+
+
+def _get_nested_metadata(metadata_key: str, metadata: dict[dict], metadata_value: str):
+    getter = metadata.get(metadata_key, {})
+    return getter.get(metadata_value, metadata_key)
+
+
+def get_agency_metadata_values(
+        df: DataFrame, 
+        agency_column: str, 
+        metadata: dict,         
+        metadata_value: str, 
+    ):
+    """Get a specific value from agency metadata (e.g., get acronym for Department of Homeland Security).
+
+    Args:
+        df (DataFrame): Input data.
+        agency_column (str): Column containing agency identifier.
+        metadata (dict): Agency metadata.
+        metadata_value (str): Value of interest from agency metadata.
+
+    Returns:
+        pd.Series: Pandas Series of new values for adding to DataFrame.
+    """    
+    if metadata_value == "acronym":
+        metadata_value = "short_name"
+    return df.loc[:, agency_column].apply(
+        lambda x: [
+            _get_nested_metadata(x_item, metadata=metadata, metadata_value=metadata_value)
+            for x_item 
+            in x
+            ] if isinstance(x, list) else _get_nested_metadata(x, metadata=metadata, metadata_value=metadata_value)
+        )

www/style.css

@@ -23,3 +23,7 @@
 #frequency {
   margin-bottom: 5% !important;
 }
+
+#menu_agency {
+  overflow: visible;
+}