PyPI - webviz-subsurface - Versions diffs - 0.2.38__tar.gz → 0.2.39__tar.gz - Mend

@@ -10,6 +10,8 @@ and this project adheres to [Semantic Versioning](https://semver.org/spec/v2.0.0
 ### Added
 - [#1297](https://github.com/equinor/webviz-subsurface/pull/1297) - Added mean for mc sensitivities in the tornadotable in `VolumetricAnalysis`.
+- [#1328](https://github.com/equinor/webviz-subsurface/pull/1328) - Use the same color for P10/P90 calculated cases for sensitivities from distributions (`TornadoPlot`)
+- [#1327](https://github.com/equinor/webviz-subsurface/pull/1327) - Added histograms to parameter distributions in `ParameterAnalysis`.
 ## [0.2.30] - 2024-06-10

@@ -1,6 +1,6 @@
 Metadata-Version: 2.1
 Name: webviz-subsurface
-Version: 0.2.38
+Version: 0.2.39
 Summary: Webviz config plugins for subsurface data
 Home-page: https://github.com/equinor/webviz-subsurface
 Author: R&T Equinor

@@ -2,7 +2,10 @@
 import pandas as pd
 import pytest
-from webviz_subsurface._components.tornado._tornado_data import TornadoData
+from webviz_subsurface._components.tornado._tornado_data import (
+    SensitivityType,
+    TornadoData,
+)
 def test_tornado_data_init():
@@ -36,6 +39,7 @@ def test_tornado_data_init():
         "values": 11.0,
         "values_ref": -26.666666666666668,
         "reals": [0],
+        "senstype": "mc",
     }
     low_high_list = tornado_data._calculate_tornado_low_high_list(avg_list)
     assert low_high_list[0] == {
@@ -52,6 +56,7 @@ def test_tornado_data_init():
         "high_tooltip": 26.666666666666668,
         "true_high": 19.0,
         "high_reals": [1],
+        "senstype": "mc",
     }
     tornado_data = TornadoData(
@@ -72,6 +77,7 @@ def test_tornado_data_init():
             "high_tooltip": 13.5,
             "true_high": 19.0,
             "high_reals": [0, 1],
+            "senstype": SensitivityType.MONTE_CARLO,
         },
         {
             "low": 0.0,
@@ -87,6 +93,7 @@ def test_tornado_data_init():
             "high_tooltip": 0.0,
             "true_high": 5.5,
             "high_reals": [2, 3],
+            "senstype": SensitivityType.SCALAR,
         },
         {
             "low": 0.0,
@@ -102,6 +109,7 @@ def test_tornado_data_init():
             "high_tooltip": 20.0,
             "true_high": 25.5,
             "high_reals": [4, 5],
+            "senstype": SensitivityType.SCALAR,
         },
         {
             "low": 0.0,
@@ -117,5 +125,6 @@ def test_tornado_data_init():
             "high_tooltip": 5.300000000000001,
             "true_high": 10.8,
             "high_reals": [6, 7],
+            "senstype": SensitivityType.MONTE_CARLO,
         },
     ]

@@ -5,7 +5,7 @@ import plotly.graph_objects as go
 from webviz_subsurface._abbreviations.number_formatting import si_prefixed
-from ._tornado_data import TornadoData
+from ._tornado_data import SensitivityType, TornadoData
 class TornadoBarChart:
@@ -136,9 +136,25 @@ class TornadoBarChart:
             hovertext.append(text)
         return hovertext
+    def get_sensitivity_colors(self, case: str) -> List:
+        """Create color list for bars based on sensitivity type
+        If colors are set by sensitivity, just create a color per sensitivty.
+        If not handle scalar and mc sensitivities separately.
+        For scalar, that is sensitivities with two "cases", use separate colors for each case.
+        For mc, use one color.
+        """
+        if self._color_by_sens:
+            return self.create_color_list(self._tornadotable["sensname"])
+        colors = []
+        for _, row in self._tornadotable.iterrows():
+            if row["senstype"] == SensitivityType.MONTE_CARLO or case == "low":
+                colors.append(self._plotly_theme["layout"]["colorway"][0])
+            else:
+                colors.append(self._plotly_theme["layout"]["colorway"][1])
+        return colors
     @property
     def data(self) -> List:
-        colors = self.create_color_list(self._tornadotable["sensname"].unique())
         return [
             {
                 "type": "bar",
@@ -157,7 +173,7 @@ class TornadoBarChart:
                 "orientation": "h",
                 "marker": {
                     "line": {"width": 1.5, "color": "black"},
-                    "color": colors if self._color_by_sens else None,
+                    "color": self.get_sensitivity_colors("low"),
                 },
             },
             {
@@ -177,7 +193,7 @@ class TornadoBarChart:
                 "orientation": "h",
                 "marker": {
                     "line": {"width": 1.5, "color": "black"},
-                    "color": colors if self._color_by_sens else None,
+                    "color": self.get_sensitivity_colors("high"),
                 },
             },
         ]
@@ -193,6 +209,15 @@ class TornadoBarChart:
     @property
     def scatter_data(self) -> List[Dict]:
+        def get_color(case_name_arr: pd.Series, case_type_arr: pd.Series) -> List:
+            colors = []
+            for case_name, case_type in zip(case_name_arr, case_type_arr):
+                if case_name == "low" or case_type == SensitivityType.MONTE_CARLO:
+                    colors.append(self._plotly_theme["layout"]["colorway"][0])
+                else:
+                    colors.append(self._plotly_theme["layout"]["colorway"][1])
+            return colors
         return [
             {
                 "type": "scatter",
@@ -202,14 +227,9 @@ class TornadoBarChart:
                 "text": df["REAL"],
                 "hovertemplate": "REAL: <b>%{text}</b><br>"
                 + "X: <b>%{x:.1f}</b> <extra></extra>",
-                "marker": {
-                    "size": 15,
-                    "color": self._plotly_theme["layout"]["colorway"][0]
-                    if case == "low"
-                    else self._plotly_theme["layout"]["colorway"][1],
-                },
+                "marker": {"size": 15, "color": get_color(df["case"], df["casetype"])},
             }
-            for case, df in self._realtable.groupby("case")
+            for _, df in self._realtable.groupby("case")
         ]
     @property

@@ -3,6 +3,15 @@ from typing import Dict, List, Optional, Union
 import numpy as np
 import pandas as pd
+from webviz_subsurface._utils.enum_shim import StrEnum
+class SensitivityType(StrEnum):
+    """Sensitivity types used in Tornado analysis."""
+    SCALAR = "scalar"
+    MONTE_CARLO = "mc"
 class TornadoData:
     REQUIRED_COLUMNS = ["REAL", "SENSNAME", "SENSCASE", "SENSTYPE", "VALUE"]
@@ -50,7 +59,11 @@ class TornadoData:
     def _create_real_df(self, dframe: pd.DataFrame) -> pd.DataFrame:
         """Make dataframe with value and case info per realization"""
         realdf = dframe[self.REQUIRED_COLUMNS].rename(
-            columns={"SENSNAME": "sensname", "SENSCASE": "senscase"}
+            columns={
+                "SENSNAME": "sensname",
+                "SENSCASE": "senscase",
+                "SENSTYPE": "senstype",
+            }
         )
         sensitivities = self._tornadotable["sensname"].unique()
@@ -61,7 +74,7 @@ class TornadoData:
                 casemask = realdf["REAL"].isin(val[f"real_{case}"])
                 realdf.loc[casemask, "case"] = case
-        mc_mask = realdf["SENSTYPE"] == "mc"
+        mc_mask = realdf["senstype"] == "mc"
         realdf["casetype"] = np.where(mc_mask, "mc", realdf["case"])
         realdf["sensname_case"] = np.where(
             mc_mask,
@@ -127,6 +140,7 @@ class TornadoData:
                                 sens_case_df["VALUE"].mean()
                             ),
                             "reals": list(map(int, sens_case_df["REAL"])),
+                            "senstype": SensitivityType.SCALAR,
                         }
                     )
             # If `SENSTYPE` is monte carlo get p10, p90
@@ -162,6 +176,7 @@ class TornadoData:
                         "values": p90,
                         "values_ref": self._scale_to_ref(p90),
                         "reals": low_reals,
+                        "senstype": SensitivityType.MONTE_CARLO,
                     }
                 )
                 avg_per_sensitivity.append(
@@ -171,6 +186,7 @@ class TornadoData:
                         "values": p10,
                         "values_ref": self._scale_to_ref(p10),
                         "reals": high_reals,
+                        "senstype": SensitivityType.MONTE_CARLO,
                     }
                 )
@@ -198,6 +214,7 @@ class TornadoData:
                     high["reals"] = []
                     high["senscase"] = None
                     high["values"] = self.reference_average
                 else:
                     low = (
                         low.copy()
@@ -218,6 +235,7 @@ class TornadoData:
                     "true_low": low["values"],
                     "low_reals": low["reals"],
                     "sensname": sensname,
+                    "senstype": sens_name_df["senstype"].unique()[0],
                     "high": self.calc_high_x(low["values_ref"], high["values_ref"]),
                     "high_base": self.calc_high_base(
                         low["values_ref"], high["values_ref"]

@@ -10,6 +10,9 @@ from webviz_config.webviz_factory import WebvizFactory
 from webviz_config.webviz_factory_registry import WEBVIZ_FACTORY_REGISTRY
 from webviz_config.webviz_instance_info import WebvizRunMode
+from webviz_subsurface._utils.design_matrix import (
+    rename_design_matrix_parameter_columns,
+)
 from webviz_subsurface._utils.perf_timer import PerfTimer
 from ..ensemble_summary_provider._arrow_unsmry_import import (
@@ -283,6 +286,7 @@ class EnsembleTableProviderFactory(WebvizFactory):
             raise ValueError(
                 f"Failed to load 'parameter.txt' files for ensemble {ens_path}."
             )
+        ensemble_df = rename_design_matrix_parameter_columns(ensemble_df)
         elapsed_load_parameters_s = timer.lap_s()

@@ -0,0 +1,36 @@
+import logging
+import pandas as pd
+LOGGER = logging.getLogger(__name__)
+def rename_design_matrix_parameter_columns(parameter_df: pd.DataFrame) -> pd.DataFrame:
+    """Given a dataframe of parameters, checks if the DESIGN_MATRIX prefix is present.
+    If present assume this is a design matrix run. Return the dataframe with the prefix
+    removed. Also do a check if removing the prefix result in any duplicates.
+    If duplicates remove those and give a warning.
+    """
+    if any(col.startswith("DESIGN_MATRIX:") for col in parameter_df.columns):
+        original_columns = parameter_df.columns
+        stripped_columns = original_columns.str.replace(
+            r"^DESIGN_MATRIX:", "", regex=True
+        )
+        rename_map = {
+            old: new
+            for old, new in zip(original_columns, stripped_columns)
+            if old != new
+        }
+        conflict_names = set(rename_map.values()) & set(original_columns)
+        if conflict_names:
+            LOGGER.info(
+                "DESIGN_MATRIX run detected, but non design matrix parameters was found."
+            )
+            LOGGER.info(
+                f"The following parameters will be dropped: {sorted(conflict_names)}"
+            )
+        parameter_df = parameter_df.drop(columns=conflict_names)
+        parameter_df = parameter_df.rename(columns=rename_map)
+    return parameter_df

@@ -2,6 +2,7 @@ from webviz_config.utils import StrEnum
 class VisualizationType(StrEnum):
+    HISTOGRAM = "histogram"
     DISTRIBUTION = "distribution"
     BOX = "box"
     STAT_TABLE = "stat-table"

@@ -8,6 +8,8 @@ from webviz_config import WebvizConfigTheme
 from webviz_subsurface._figures import create_figure
 from webviz_subsurface._models.parameter_model import ParametersModel as Pmodel
+from .._types import VisualizationType
 class ParametersModel:
     """Class to process and visualize ensemble parameter data"""
@@ -163,7 +165,7 @@ class ParametersModel:
         self,
         ensembles: list,
         parameters: List[Any],
-        plot_type: str = "distribution",
+        plot_type: VisualizationType = VisualizationType.DISTRIBUTION,
     ) -> go.Figure:
         """Create subplots for selected parameters"""
         df = self.dataframe_melted.copy()
@@ -171,7 +173,7 @@ class ParametersModel:
         df = df[df["PARAMETER"].isin(parameters)]
         df = self._sort_parameters_col(df, parameters)
-        return (
+        figure = (
             create_figure(
                 plot_type=plot_type,
                 data_frame=df,
@@ -179,6 +181,7 @@ class ParametersModel:
                 facet_col="PARAMETER",
                 color="ENSEMBLE",
                 color_discrete_sequence=self.colorway,
+                barmode="overlay",
             )
             .update_xaxes(matches=None)
             .for_each_trace(
@@ -189,6 +192,11 @@ class ParametersModel:
                 )
             )
         )
+        # Use bingroup=None so that Plotly calculates bins per trace
+        # This also means that individual ensembles will have separate binning.
+        if plot_type == VisualizationType.HISTOGRAM:
+            figure.update_traces(bingroup=None)
+        return figure
     def get_stat_value(self, parameter: str, ensemble: str, stat_column: str) -> float:
         """

@@ -20,6 +20,7 @@ class ParamDistVisualizationType(SettingsGroupABC):
             wcc.RadioItems(
                 id=self.register_component_unique_id(self.Ids.VISUALIZATION_TYPE),
                 options=[
+                    {"label": "Histogram", "value": VisualizationType.HISTOGRAM},
                     {
                         "label": "Distribution plots",
                         "value": VisualizationType.DISTRIBUTION,
@@ -30,7 +31,7 @@ class ParamDistVisualizationType(SettingsGroupABC):
                         "value": VisualizationType.STAT_TABLE,
                     },
                 ],
-                value=VisualizationType.DISTRIBUTION,
+                value=VisualizationType.HISTOGRAM,
                 vertical=True,
             )
         ]

@@ -214,6 +214,7 @@ webviz_subsurface/_providers/well_provider/well_server.py
 webviz_subsurface/_utils/__init__.py
 webviz_subsurface/_utils/colors.py
 webviz_subsurface/_utils/dataframe_utils.py
+webviz_subsurface/_utils/design_matrix.py
 webviz_subsurface/_utils/ensemble_summary_provider_set.py
 webviz_subsurface/_utils/ensemble_summary_provider_set_factory.py
 webviz_subsurface/_utils/ensemble_table_provider_set.py

webviz-subsurface 0.2.38__tar.gz → 0.2.39__tar.gz

webviz-subsurface 0.2.38tar.gz → 0.2.39tar.gz