fix: use underscores instead of hyphens in simulation table column names

docs/user-guide/outputs.md

@@ -56,7 +56,7 @@ df = results.data
 ~~~
 
 The DataFrame has columns: `block`, `component`, `output`,
-`absolute-time-index`, `block-time-index`, `scenario-index`, `value`, `basis-status`.
+`absolute_time_index`, `block_time_index`, `scenario_index`, `value`, `basis_status`.
 
 Reading the value of the optimisation variable `var_id` of component `component_id`
 for a single time step and scenario:
@@ -65,12 +65,12 @@ for a single time step and scenario:
 value = df[(df["component"] == component_id) & (df["output"] == var_id)]["value"].iloc[0]
 ~~~
 
-For multi-time or multi-scenario results, filter additionally by `block-time-index`
-and `scenario-index`:
+For multi-time or multi-scenario results, filter additionally by `block_time_index`
+and `scenario_index`:
 
 ~~~ python
 sub = df[(df["component"] == component_id) & (df["output"] == var_id)]
-value_s0_t1 = sub[(sub["scenario-index"] == 0) & (sub["block-time-index"] == 1)]["value"].iloc[0]
+value_s0_t1 = sub[(sub["scenario_index"] == 0) & (sub["block_time_index"] == 1)]["value"].iloc[0]
 ~~~
 
 ---

src/gems/simulation/simulation_table.py

@@ -15,7 +15,7 @@ class OutputView:
     """
 
     def __init__(self, df: pd.DataFrame) -> None:
-        # df: index = absolute-time-index, columns = scenario-index
+        # df: index = absolute_time_index, columns = scenario_index
         self._df = df
 
     @property
@@ -32,8 +32,8 @@ def value(
 
         Called with no arguments returns the full Time × Scenario DataFrame.
         Called with one argument returns a ``pd.Series``:
-        - ``value(scenario_index=s)`` → Series indexed by absolute-time-index
-        - ``value(time_index=t)``     → Series indexed by scenario-index
+        - ``value(scenario_index=s)`` → Series indexed by absolute_time_index
+        - ``value(time_index=t)``     → Series indexed by scenario_index
         Called with both arguments returns a scalar ``float``.
         """
         if time_index is None and scenario_index is None:
@@ -142,7 +142,7 @@ def to_dataset(self) -> xr.Dataset:
         """Return simulation results as an xr.Dataset.
 
         Each output variable becomes a DataArray with dimensions
-        (component, absolute-time-index, scenario-index).
+        (component, absolute_time_index, scenario_index).
         Scalar rows without component/time/scenario (e.g. objective-value)
         are stored as zero-dimensional variables.
         """
@@ -174,11 +174,11 @@ class SimulationColumns(str, Enum):
     BLOCK = "block"
     COMPONENT = "component"
     OUTPUT = "output"
-    ABSOLUTE_TIME_INDEX = "absolute-time-index"
-    BLOCK_TIME_INDEX = "block-time-index"
-    SCENARIO_INDEX = "scenario-index"
+    ABSOLUTE_TIME_INDEX = "absolute_time_index"
+    BLOCK_TIME_INDEX = "block_time_index"
+    SCENARIO_INDEX = "scenario_index"
     VALUE = "value"
-    BASIS_STATUS = "basis-status"
+    BASIS_STATUS = "basis_status"
 
 
 class SimulationTableBuilder:
@@ -441,7 +441,7 @@ def _da_to_df(
     ) -> pd.DataFrame:
         """Vectorize a [component?, time?, scenario?] DataArray into a DataFrame.
 
-        Index columns (absolute-time-index, block-time-index, scenario-index) are
+        Index columns (absolute_time_index, block_time_index, scenario_index) are
         set to None for dimensions that are absent from the original DataArray,
         signalling that the output is independent of that dimension.
         """

tests/e2e/functional/test_optim_modes.py

@@ -83,8 +83,8 @@
 _KEY_COLS = [
     "component",
     "output",
-    "absolute-time-index",
-    "scenario-index",
+    "absolute_time_index",
+    "scenario_index",
 ]
 
 

tests/e2e/functional/test_rolling_horizon_suboptimality.py

@@ -120,7 +120,7 @@ def _get_value(raw, component: str, output: str, timestep: int) -> float:
     mask = (
         (raw["component"] == component)
         & (raw["output"] == output)
-        & (raw["absolute-time-index"] == timestep)
+        & (raw["absolute_time_index"] == timestep)
     )
     rows = raw[mask]
     assert (

tests/e2e/functional/test_simtable_timeblock.py

@@ -18,7 +18,7 @@
   - System: 3 components — node, generator (pmax=200), demand (timevarying: demand[t]=t)
   - Data horizon: 150 timesteps
   - TimeBlock: [40, 90)  →  50 timesteps (indices 40–89)
-  - Checks that SimulationTable absolute-time-index, block-time-index, and
+  - Checks that SimulationTable absolute_time_index, block_time_index, and
     generation values are all consistent with the partial block.
 """
 
@@ -52,8 +52,8 @@ def test_simtable_on_partial_timeblock(lib_dict_unittest: dict[str, Library]) ->
     With demand[t] = t and pmax = 200, the optimizer sets generation[t] = t at
     every timestep, making all three assertions self-consistent:
       1. generation values equal their absolute timestep index.
-      2. absolute-time-index runs from 40 to 89 (not 0–149 or 0–49).
-      3. block-time-index runs from 0 to 49 and equals absolute-time-index − 40.
+      2. absolute_time_index runs from 40 to 89 (not 0–149 or 0–49).
+      3. block_time_index runs from 0 to 49 and equals absolute_time_index − 40.
     """
     node_model = lib_dict_unittest["basic"].models["basic.node"]
     generator_model = lib_dict_unittest["basic"].models["basic.generator"]
@@ -120,7 +120,7 @@ def test_simtable_on_partial_timeblock(lib_dict_unittest: dict[str, Library]) ->
     assert abs_times == list(range(BLOCK_START, BLOCK_END))
     assert block_times == list(range(BLOCK_END - BLOCK_START))
 
-    # absolute-time-index = block-time-index + BLOCK_START for every row
+    # absolute_time_index = block_time_index + BLOCK_START for every row
     offset = (
         gen_rows[SimulationColumns.ABSOLUTE_TIME_INDEX.value].astype(int)
         - gen_rows[SimulationColumns.BLOCK_TIME_INDEX.value].astype(int)

tests/unittests/simulation/test_simulation_table_export.py

@@ -126,7 +126,7 @@ def test_to_dataset_values_match_data_single_scenario() -> None:
         expected = float(row[SimulationColumns.VALUE.value])
         actual = float(
             ds[output].sel(
-                component=comp, **{"absolute-time-index": t, "scenario-index": s}
+                component=comp, **{"absolute_time_index": t, "scenario_index": s}
             )
         )
         assert actual == pytest.approx(expected)
@@ -144,7 +144,7 @@ def test_to_dataset_values_match_data_multi_scenario() -> None:
         expected = float(row[SimulationColumns.VALUE.value])
         actual = float(
             ds[output].sel(
-                component=comp, **{"absolute-time-index": t, "scenario-index": s}
+                component=comp, **{"absolute_time_index": t, "scenario_index": s}
             )
         )
         assert actual == pytest.approx(expected)