Rename the timeslices table's timeslice_id column to timeslice

src/ispypsa/pypsa_build/links.py

@@ -42,7 +42,7 @@ def _add_links_to_network(
             CQ-NQ_existing  p_min_pu   ,                -0.714  # fallback only
 
         timeslice_snapshots:
-            timeslice_id     investment_periods  snapshots
+            timeslice        investment_periods  snapshots
             qld_peak_demand  2025                2025-01-13 12:00
 
         network.snapshots:
@@ -115,7 +115,7 @@ def _build_link_pu_overrides(
             CQ-NQ_existing  p_min_pu   ,                -0.714  # fallback only
 
         timeslice_snapshots:
-            timeslice_id     investment_periods  snapshots
+            timeslice        investment_periods  snapshots
             qld_peak_demand  2025                2025-01-13 12:00
 
         snapshots:
@@ -161,7 +161,7 @@ def _expand_limits_to_snapshots(
             CQ-NQ_existing  p_min_pu   qld_peak_demand  -0.9    # no fallback
 
         timeslice_snapshots:
-            timeslice_id     investment_periods  snapshots
+            timeslice        investment_periods  snapshots
             qld_peak_demand  2025                2025-01-13 12:00
 
         snapshots:
@@ -234,14 +234,14 @@ def _place_named_limits_at_snapshots(
             CQ-NQ_existing  p_max_pu   qld_peak_demand  0.857
 
         timeslice_snapshots:
-            timeslice_id     investment_periods  snapshots
+            timeslice        investment_periods  snapshots
             qld_peak_demand  2025                2025-01-13 12:00
 
         returns:
             name            attribute  investment_periods  snapshots         value
             CQ-NQ_existing  p_max_pu   2025                2025-01-13 12:00  0.857
     """
-    active_at = timeslice_snapshots.rename(columns={"timeslice_id": "timeslice"})
+    active_at = timeslice_snapshots.copy()
     active_at["snapshots"] = pd.to_datetime(active_at["snapshots"])
     placed = named.merge(active_at, on="timeslice")
     return placed.loc[:, _LIMIT_PER_SNAPSHOT_COLUMNS]

src/ispypsa/templater/timeslices.py

@@ -21,7 +21,7 @@
 from ispypsa.templater.custom_constraints_from_plexos import _tag_to_timeslice
 
 _TIMESLICE_COLUMNS = [
-    "timeslice_id",
+    "timeslice",
     "reference_year",
     "start_month_day",
     "end_month_day",
@@ -90,7 +90,7 @@ def _template_timeslices(
             2026           2015
 
         returns:
-            timeslice_id     reference_year  start_month_day  end_month_day
+            timeslice        reference_year  start_month_day  end_month_day
             nsw_peak_demand  2015            11-18            11-20  # end exclusive
     """
     events = _parse_calendar_events(timeslice_calendar)
@@ -115,29 +115,29 @@ def _parse_calendar_events(calendar: pd.DataFrame) -> pd.DataFrame:
 
     I/O Example:
         DATETIME=18/11/2021, NAME="NSW Hot Day", TIMESLICE=-1
-        -> DATETIME=2021-11-18, timeslice_id="nsw_peak_demand", TIMESLICE=-1
+        -> DATETIME=2021-11-18, timeslice="nsw_peak_demand", TIMESLICE=-1
     """
     events = calendar.copy()
     events["DATETIME"] = pd.to_datetime(events["DATETIME"], dayfirst=True)
-    events["timeslice_id"] = events["NAME"].map(_tag_to_timeslice)
-    return events.sort_values(["timeslice_id", "DATETIME"])
+    events["timeslice"] = events["NAME"].map(_tag_to_timeslice)
+    return events.sort_values(["timeslice", "DATETIME"])
 
 
 def _add_next_event_columns(events: pd.DataFrame) -> pd.DataFrame:
     """Annotates each event with the date and state of the next event for the
     same timeslice (NaN/NaT on each timeslice's last event).
 
     I/O Example:
-        timeslice_id     DATETIME    TIMESLICE
+        timeslice        DATETIME    TIMESLICE
         nsw_peak_demand  2021-11-18  -1
         nsw_peak_demand  2021-11-20  0
 
         ->
-        timeslice_id     DATETIME    TIMESLICE  next_date   next_state
+        timeslice        DATETIME    TIMESLICE  next_date   next_state
         nsw_peak_demand  2021-11-18  -1         2021-11-20  0
         nsw_peak_demand  2021-11-20  0          NaT         NaN
     """
-    grouped = events.groupby("timeslice_id")
+    grouped = events.groupby("timeslice")
     events["next_date"] = grouped["DATETIME"].shift(-1)
     events["next_state"] = grouped["TIMESLICE"].shift(-1)
     return events
@@ -163,19 +163,19 @@ def _extract_windows(events: pd.DataFrame) -> pd.DataFrame:
     """Turns each on event into a window row ending at the paired off event.
 
     I/O Example:
-        timeslice_id     DATETIME    TIMESLICE  next_date   next_state
+        timeslice        DATETIME    TIMESLICE  next_date   next_state
         nsw_peak_demand  2021-11-18  -1         2021-11-20  0
         nsw_peak_demand  2021-11-20  0          NaT         NaN
 
         returns:
-            timeslice_id     start_date  end_date
+            timeslice        start_date  end_date
             nsw_peak_demand  2021-11-18  2021-11-20
     """
     windows = events[events["TIMESLICE"] == -1]
     windows = windows.rename(
         columns={"DATETIME": "start_date", "next_date": "end_date"}
     )
-    return windows[["timeslice_id", "start_date", "end_date"]].reset_index(drop=True)
+    return windows[["timeslice", "start_date", "end_date"]].reset_index(drop=True)
 
 
 def _drop_horizon_truncated_planning_years(windows: pd.DataFrame) -> pd.DataFrame:
@@ -187,7 +187,7 @@ def _drop_horizon_truncated_planning_years(windows: pd.DataFrame) -> pd.DataFram
     nothing is lost by dropping the year entirely.
 
     I/O Example:
-        timeslice_id          start_date  end_date    planning_year
+        timeslice             start_date  end_date    planning_year
         nsw_peak_demand       2057-11-18  2057-11-20  2058           # dropped: shares
         nsw_winter_reference  2058-04-01  NaT         2058           # the truncated year
         nsw_peak_demand       2056-11-18  2056-11-20  2057           # kept
@@ -228,7 +228,7 @@ def _extend_sequence_to_horizon(
             2027           2011
 
         windows (only planning_year is read):
-            timeslice_id     start_date  end_date    planning_year
+            timeslice        start_date  end_date    planning_year
             nsw_peak_demand  2028-11-18  2028-11-20  2029
 
         returns:
@@ -256,17 +256,17 @@ def _convert_windows_to_month_days(windows: pd.DataFrame) -> pd.DataFrame:
     next calendar year, and winter's 04-01 -> 10-01 extends past 30 June).
 
     I/O Example:
-        timeslice_id     start_date  end_date    planning_year  reference_year
+        timeslice        start_date  end_date    planning_year  reference_year
         nsw_peak_demand  2025-11-18  2025-11-20  2026           2015
 
         ->
-        timeslice_id     reference_year  planning_year  start_month_day  end_month_day
+        timeslice        reference_year  planning_year  start_month_day  end_month_day
         nsw_peak_demand  2015            2026           11-18            11-20
     """
     windows["start_month_day"] = windows["start_date"].dt.strftime("%m-%d")
     windows["end_month_day"] = windows["end_date"].dt.strftime("%m-%d")
     return windows[
-        ["timeslice_id", "reference_year", "planning_year"]
+        ["timeslice", "reference_year", "planning_year"]
         + ["start_month_day", "end_month_day"]
     ]
 
@@ -278,7 +278,7 @@ def _raise_on_inconsistent_reference_year_patterns(patterns: pd.DataFrame) -> No
     decoding one pattern per reference year would silently lose windows."""
     occurrences = patterns.groupby(["reference_year", "planning_year"]).apply(
         lambda x: frozenset(
-            zip(x["timeslice_id"], x["start_month_day"], x["end_month_day"])
+            zip(x["timeslice"], x["start_month_day"], x["end_month_day"])
         ),
         include_groups=False,
     )
@@ -296,12 +296,12 @@ def _keep_first_occurrence_per_reference_year(patterns: pd.DataFrame) -> pd.Data
     occurrences are identical — validated before this is called).
 
     I/O Example:
-        timeslice_id     reference_year  planning_year  start_month_day  end_month_day
+        timeslice        reference_year  planning_year  start_month_day  end_month_day
         nsw_peak_demand  2015            2026           11-18            11-20
         nsw_peak_demand  2015            2031           11-18            11-20
 
         returns:
-            timeslice_id     reference_year  start_month_day  end_month_day
+            timeslice        reference_year  start_month_day  end_month_day
             nsw_peak_demand  2015            11-18            11-20
     """
     first_occurrence = patterns.groupby("reference_year")["planning_year"].transform(
@@ -316,12 +316,12 @@ def _raise_unless_windows_tile_the_year(timeslices: pd.DataFrame) -> None:
     """Raise unless, within every region and reference year, the windows tile
     the year exactly — no day left uncovered and none covered twice. A gap there
     would let a snapshot fall in no timeslice (silently taking a base limit); an
-    overlap would let it fall in two. The region is the timeslice_id prefix
+    overlap would let it fall in two. The region is the timeslice prefix
     before the first underscore.
     """
-    # add a region column from the timeslice_id prefix (nsw_peak_demand -> nsw)
+    # add a region column from the timeslice prefix (nsw_peak_demand -> nsw)
     # to group by, so each region's windows are checked for tiling independently
-    tagged = timeslices.assign(region=timeslices["timeslice_id"].str.split("_").str[0])
+    tagged = timeslices.assign(region=timeslices["timeslice"].str.split("_").str[0])
     not_tiling = sorted(
         (region, int(year))
         for (region, year), windows in tagged.groupby(["region", "reference_year"])
@@ -371,8 +371,8 @@ def _raise_unless_only_winter_crosses_financial_year(timeslices: pd.DataFrame) -
         ),
         axis=1,
     )
-    is_winter = timeslices["timeslice_id"].str.endswith("_winter_reference")
-    crossing = sorted(timeslices.loc[spans_july & ~is_winter, "timeslice_id"].unique())
+    is_winter = timeslices["timeslice"].str.endswith("_winter_reference")
+    crossing = sorted(timeslices.loc[spans_july & ~is_winter, "timeslice"].unique())
     if crossing:
         raise ValueError(
             f"Only winter_reference windows may cross the 1 July financial-year "
@@ -403,8 +403,8 @@ def _raise_unless_winter_is_constant_per_region(timeslices: pd.DataFrame) -> Non
     reference years, so which reference year owns its end never changes its
     value. This is the assumption documented on _template_timeslices.
     """
-    winter = timeslices[timeslices["timeslice_id"].str.endswith("_winter_reference")]
-    winter = winter.assign(region=winter["timeslice_id"].str.split("_").str[0])
+    winter = timeslices[timeslices["timeslice"].str.endswith("_winter_reference")]
+    winter = winter.assign(region=winter["timeslice"].str.split("_").str[0])
     varying = sorted(
         region
         for region, group in winter.groupby("region")

src/ispypsa/translator/timeslices.py

@@ -4,7 +4,7 @@
 
 logger = logging.getLogger(__name__)
 
-_TIMESLICE_SNAPSHOT_COLUMNS = ["timeslice_id", "investment_periods", "snapshots"]
+_TIMESLICE_SNAPSHOT_COLUMNS = ["timeslice", "investment_periods", "snapshots"]
 
 
 def _create_timeslice_snapshot_mapping(
@@ -20,14 +20,14 @@ def _create_timeslice_snapshot_mapping(
     Mapping is done on a model year by model year basis:
         - Each model year is mapped to a reference year.
         - Then the timeslice windows for that reference year are used to map each
-          snapshot to a timeslice_id.
+          snapshot to a timeslice.
 
     I/O Example:
         timeslices (each year's windows tile (cover in full) its financial year:
         summer opens in November and wraps past New Year, a peak day interrupts it,
         summer resumes to April, then winter runs April to November, crossing 30
         June into the next financial year):
-            timeslice_id          reference_year  start_month_day  end_month_day
+            timeslice             reference_year  start_month_day  end_month_day
             nsw_summer_typical    2011            11-01            01-31
             nsw_peak_demand       2011            01-31            02-02
             nsw_summer_typical    2011            02-02            04-01
@@ -49,7 +49,7 @@ def _create_timeslice_snapshot_mapping(
             -> FY2025 uses 2011's pattern, FY2026 uses 2018's
 
         returns:
-            timeslice_id          investment_periods  snapshots
+            timeslice             investment_periods  snapshots
             nsw_winter_reference  2025                2024-08-15 12:00:00
             nsw_summer_typical    2025                2025-01-20 12:00:00
             nsw_peak_demand       2025                2025-01-31 12:00:00
@@ -111,14 +111,14 @@ def _tag_snapshots_with_pattern(
             2026                2026-01-07 12:00:00  01-07
 
         pattern (reference year 2018):
-            timeslice_id          start_month_day  end_month_day
+            timeslice             start_month_day  end_month_day
             nsw_summer_typical    11-01            01-07
             nsw_peak_demand       01-07            01-08
             nsw_summer_typical    01-08            04-01
             nsw_winter_reference  04-01            11-01
 
         returns:
-            timeslice_id          investment_periods  snapshots
+            timeslice             investment_periods  snapshots
             nsw_winter_reference  2026                2025-08-15 12:00:00
             nsw_peak_demand       2026                2026-01-07 12:00:00
     """
@@ -154,7 +154,7 @@ def _concat_tagged_snapshots(mapped: list[pd.DataFrame]) -> pd.DataFrame:
     """Combines the per-model-year tagged snapshots into one mapping table,
     in snapshot order."""
     mapping = pd.concat(mapped, ignore_index=True)
-    mapping = mapping.sort_values(["snapshots", "timeslice_id"]).reset_index(drop=True)
+    mapping = mapping.sort_values(["snapshots", "timeslice"]).reset_index(drop=True)
     return mapping.loc[:, _TIMESLICE_SNAPSHOT_COLUMNS]
 
 
@@ -174,7 +174,7 @@ def _log_referenced_timeslices_without_snapshots(
     referenced = set(link_timeslice_limits["timeslice"]) | set(
         custom_constraints_rhs["timeslice"].dropna()
     )
-    without_snapshots = referenced - set(timeslice_snapshots["timeslice_id"])
+    without_snapshots = referenced - set(timeslice_snapshots["timeslice"])
     if without_snapshots:
         logger.warning(
             f"Timeslices referenced by transmission limits or custom constraints "

src/ispypsa/validation/schemas/custom_constraints_rhs.yaml

@@ -22,7 +22,7 @@ columns:
     type: string
     required: false
     allowed_values_from:
-      - timeslices: timeslice_id
+      - timeslices: timeslice
     description: >
       Demand condition the limit applies to: the constraint binds only at
       snapshots inside the timeslice's active windows.

src/ispypsa/validation/schemas/network_transmission_path_limits.yaml

@@ -33,7 +33,7 @@ custom_validation:
       A path that names timeslices must name them all for its region, or add an empty-timeslice
       fallback, in both directions; a single all-empty row covers the whole space.
       Because the timeslices table's windows partition the year, covering every
-      region timeslice_id covers every snapshot downstream.
+      region timeslice covers every snapshot downstream.
 columns:
   path_id:
     type: string
@@ -60,10 +60,10 @@ columns:
     type: string
     required: false
     allowed_values_from:
-      - timeslices: timeslice_id
+      - timeslices: timeslice
     description: >
       Demand condition the limit applies to: the limit binds only at snapshots
-      inside the timeslice's active windows. A timeslice_id's region is the prefix
+      inside the timeslice's active windows. A timeslice's region is the prefix
       before its first underscore (e.g. qld in qld_peak_demand).
 
       If absent (or empty):

src/ispypsa/validation/schemas/timeslices.yaml

@@ -1,17 +1,17 @@
 table: timeslices
 required: false
 unique:
-  - [timeslice_id, reference_year, start_month_day]
+  - [timeslice, reference_year, start_month_day]
 custom_validation:
   - name: no_overlapping_windows_per_region_and_reference_year
     description: >
-      Within a region (the timeslice_id prefix before the first underscore)
+      Within a region (the timeslice prefix before the first underscore)
       and reference_year, no two windows may cover the same day. Windows are
       compared as [start, end) month-day ranges where an end_month_day at or
       before its start wraps past New Year (e.g. 11-20 to 03-20).
   - name: windows_cover_full_year_per_region_and_reference_year
     description: >
-      Within a region (the timeslice_id prefix before the first underscore)
+      Within a region (the timeslice prefix before the first underscore)
       and reference_year, the windows must cover every day of the year, leaving
       no day with no active timeslice. Together with
       no_overlapping_windows_per_region_and_reference_year this makes the
@@ -62,7 +62,7 @@ description: >
   No timeslice is ever active: timeslice-tagged transmission path limits and
   custom-constraint RHS values never apply to any snapshot.
 columns:
-  timeslice_id:
+  timeslice:
     type: string
     required: true
     allowed_values: [

tests/test_cli/test_create_ispypsa_inputs_new_table_formats.py

@@ -454,7 +454,7 @@ def test_create_ispypsa_inputs_new_format(
     # RefYear5000 calendar, identical at every granularity.
     timeslices = pd.read_csv(output_dir / "timeslices.csv")
     assert len(timeslices) == _NUM_TIMESLICE_PATTERN_ROWS_75
-    assert timeslices["timeslice_id"].nunique() == _NUM_TIMESLICE_IDS_75
+    assert timeslices["timeslice"].nunique() == _NUM_TIMESLICE_IDS_75
     assert timeslices["reference_year"].nunique() == _NUM_REFERENCE_YEARS_75
 
     # custom_constraints — populated at sub_regions, header-only elsewhere (see