51 1d resample fix (#53)

* Fixes #51 - 1D resampling fix for custom mass names

* improved check before solve

Author: elphick

CHANGELOG.rst

@@ -1,3 +1,12 @@
+Geomet 0.4.11 (2025-01-25)
+==========================
+
+Bugfix
+------
+
+- 1D resampling fix for custom mass names (#51)
+
+
 Geomet 0.4.10 (2025-01-23)
 ==========================
 

elphick/geomet/flowsheet/flowsheet.py

@@ -15,6 +15,7 @@
 import yaml
 from matplotlib import pyplot as plt
 from matplotlib.colors import ListedColormap, LinearSegmentedColormap
+from networkx.algorithms.dag import is_directed_acyclic_graph
 from plotly.subplots import make_subplots
 
 from elphick.geomet import Sample
@@ -306,6 +307,8 @@ def copy_without_stream_data(self):
     def solve(self):
         """Solve missing streams"""
 
+        assert is_directed_acyclic_graph(self.graph), "Graph is not a Directed Acyclic Graph (DAG), so cannot be solved."
+
         # Check the number of missing mc's on edges in the network
         missing_count: int = sum([1 for u, v, d in self.graph.edges(data=True) if d['mc'] is None])
         prev_missing_count = missing_count + 1  # Initialize with a value greater than missing_count

elphick/geomet/interval_sample.py

@@ -585,6 +585,7 @@ def resample_1d(self, interval_edges: Union[Iterable, int],
         df_upsampled: pd.DataFrame = mass_preserving_interp(self.mass_data,
                                                             interval_edges=interval_edges, precision=precision,
                                                             include_original_edges=include_original_edges,
+                                                            mass_wet=None, mass_dry=self.mass_dry_var,
                                                             interval_data_as_mass=True)
 
         obj: IntervalSample = IntervalSample(df_upsampled, name=self.name, moisture_in_scope=False,

elphick/geomet/utils/interp.py

@@ -9,7 +9,7 @@
 
 def mass_preserving_interp(df_intervals: pd.DataFrame, interval_edges: Union[Iterable, int],
                            include_original_edges: bool = True, precision: Optional[int] = None,
-                           mass_wet: str = 'mass_wet', mass_dry: str = 'mass_dry',
+                           mass_wet: Optional[str] = 'mass_wet', mass_dry: str = 'mass_dry',
                            interval_data_as_mass: bool = False) -> pd.DataFrame:
     """Interpolate with zero mass loss using pchip
 
@@ -81,7 +81,7 @@ def mass_preserving_interp(df_intervals: pd.DataFrame, interval_edges: Union[Ite
     # if the new grid extrapolates (on the coarse side), mass will be lost, so we assume that when extrapolating.
     # the mass in the extrapolated fractions is zero.  By inserting these records the spline will conform.
     x_extra = grid_vals[grid_vals > mass_cum.index.max()]
-    if x_extra:
+    if len(x_extra) > 0:
         cum_max: pd.Series = mass_cum.iloc[-1, :]
         mass_cum = mass_cum.reindex(index=mass_cum.index.append(pd.Index(x_extra)))  # reindex to enable insert
         mass_cum.loc[x_extra, :] = cum_max.values
@@ -180,7 +180,8 @@ def mass_preserving_interp_2d(intervals: pd.DataFrame, interval_edges: dict[str,
             # proportion the dim_1 interpolated values by the dim_2 recovery
             new_vals: pd.DataFrame = dim_2_deportment.mul(filtered_intervals_mass.loc[dim_1_interp_interval].values)
             # create a multiindex by combining the dim_1 and dim_2 intervals
-            new_index = pd.MultiIndex.from_arrays([pd.IntervalIndex([dim_1_interp_interval] * len(new_vals)), new_vals.index])
+            new_index = pd.MultiIndex.from_arrays(
+                [pd.IntervalIndex([dim_1_interp_interval] * len(new_vals)), new_vals.index])
             new_vals.index = new_index
             chunks.append(new_vals)
 
@@ -190,4 +191,3 @@ def mass_preserving_interp_2d(intervals: pd.DataFrame, interval_edges: dict[str,
     # convert to composition
     res = mass_to_composition(res, mass_dry=mass_dry)
     return res
-

elphick/geomet/utils/pandas.py

@@ -88,7 +88,7 @@ def composition_to_mass(df: pd.DataFrame,
         df: The pd.DataFrame containing mass.  H2O if provided will be ignored.  All columns other than the
          mass_wet and mass_dry are assumed to be `additive`, that is, dry mass weighting is valid.
          Assumes composition is in %w/w units.
-        mass_wet: The wet mass column, optional. If not provided, it's assumed to be equal to mass_dry.
+        mass_wet: The wet mass column, optional.
         mass_dry: The dry mass column, not optional.  Consider solve_mass_moisture prior to this call if needed.
         moisture_column_name: if mass_wet is provided, the resultant moisture will be returned with this column name.
          If None, and moisture is detected in the input, that column name will be used instead.

pyproject.toml

@@ -1,7 +1,7 @@
 [tool.poetry]
 name = "geometallurgy"
 packages = [{ include = "elphick/geomet" }]
-version = "0.4.10"
+version = "0.4.11"
 description = "Tools for the geometallurgist"
 authors = ["Greg <11791585+elphick@users.noreply.github.com>"]
 repository = "https://github.com/elphick/geometallurgy"