Add two-output graph for visualization.py. Add csv validation for dashboard to catch wrong delimiters or incorrect column names. Closes #45. Closes #50.

Author: justushelo

panel/simdec_app.py

@@ -1,5 +1,6 @@
 import bisect
 import io
+import re
 
 from bokeh.models import PrintfTickFormatter
 from bokeh.models.widgets.tables import NumberFormatter
@@ -15,10 +16,8 @@
 from simdec.sensitivity_indices import SensitivityAnalysisResult
 from simdec.visualization import sequential_cmaps, single_color_to_colormap
 
-
 # panel app
-pn.extension("tabulator")
-pn.extension("floatpanel")
+pn.extension("tabulator", "floatpanel", notifications=True)
 
 pn.config.sizing_mode = "stretch_width"
 pn.config.throttled = True
@@ -43,34 +42,82 @@
 )
 
 
+def _validate_csv_bytes(raw_bytes):
+    """Pre-parse validation. Returns an error string or None."""
+    try:
+        first_line = raw_bytes.decode("utf-8").split("\n")[0].strip()
+    except UnicodeDecodeError:
+        return "File encoding error. Please use files in UTF-8."
+
+    if "," not in first_line:
+        detected = (
+            "Semicolons(';')"
+            if ";" in first_line
+            else "tabs"
+            if "\t" in first_line
+            else "Unknown delimiter"
+        )
+        return f"Wrong column delimiter {detected}. Save the data with commas ',' as the delimiter"
+
+    col_names = [c.strip().strip('"').strip("'") for c in first_line.split(",")]
+    bad_cols = [c for c in col_names if re.search(r"[^A-Za-z0-9_ \-.]", c)]
+    if bad_cols:
+        return (
+            f"Special characters found in column name(s): {bad_cols}."
+            f"Column names may contain only letters, numbers and underscores."
+            f"Please rename columns {bad_cols} before uploading data again."
+        )
+    return None
+
+
 @pn.cache
 def load_data(text_fname):
     if text_fname is None:
-        text_fname = "tests/data/stress.csv"
-    else:
-        text_fname = io.BytesIO(text_fname)
+        return pd.read_csv("tests/data/stress.csv")
+
+    raw_bytes = bytes(text_fname)
 
-    data = pd.read_csv(text_fname)
-    return data
+    # Run pre-validation
+    error = _validate_csv_bytes(raw_bytes)
+    if error:
+        pn.state.notifications.error(error, duration=0)
+        return None
+
+    # Try parsing
+    try:
+        text_fname = io.BytesIO(text_fname)
+        return pd.read_csv(text_fname)
+    except Exception as e:
+        pn.state.notifications.error(f"Could not parse CSV {e}.", duration=0)
+        return None
 
 
 @pn.cache
 def column_inputs(data, output):
+    if data is None:
+        return []
     inputs = list(data.columns)
-    inputs.remove(output)
+    if output in inputs:
+        inputs.remove(output)
     return inputs
 
 
 @pn.cache
 def column_output(data):
+    if data is None:
+        return []
     return list(data.columns)
 
 
 @pn.cache
 def filtered_data(data, output_name):
+    if data is None or not output_name:
+        return pd.Series(dtype=float)
     try:
         return data[output_name]
     except KeyError:
+        if isinstance(output_name, list):
+            return data.iloc[:, [0]]
         return data.iloc[:, 0]
 
 
@@ -350,7 +397,7 @@ def csv_data(
 
 interactive_column_output = pn.bind(column_output, interactive_file)
 # hack to make the default selection faster
-interactive_output_ = pn.bind(lambda x: x[0], interactive_column_output)
+interactive_output_ = pn.bind(lambda x: x[0] if x else None, interactive_column_output)
 selector_output = pn.widgets.Select(
     name="Output", value=interactive_output_, options=interactive_column_output
 )

src/simdec/visualization.py

@@ -1,7 +1,7 @@
 import copy
 import functools
 import itertools
-from typing import Literal
+from typing import Literal, Optional
 
 import colorsys
 import matplotlib as mpl
@@ -135,17 +135,25 @@ def palette(
 def visualization(
     *,
     bins: pd.DataFrame,
+    bins2: Optional[pd.DataFrame] = None,
     palette: list[list[float]],
     n_bins: str | int = "auto",
     kind: Literal["histogram", "boxplot"] = "histogram",
     ax=None,
+    output_name: str = "Output 1",
+    output_name2: str = "Output 2",
+    xlim: Optional[tuple[float, float]] = None,
+    ylim: Optional[tuple[float, float]] = None,
+    r_scatter: float = 1.0,
 ) -> plt.Axes:
     """Histogram plot of scenarios.
 
     Parameters
     ----------
     bins : DataFrame
         Multidimensional bins.
+    bins2 : DataFrame
+        Multidimensional bins for output 2
     palette : list of int of size (n, 4)
         List of colours corresponding to scenarios.
     n_bins : str or int
@@ -154,16 +162,85 @@ def visualization(
         Histogram or Box Plot.
     ax : Axes, optional
         Matplotlib axis.
+    output_name : str, default "Output 1"
+        Name of the primary output variable.
+    output_name2 : str, default "Output 2"
+        Name of the second output variable.
+    xlim : tuple of float, optional
+        Minimum and maximum values for the x-axis (Output 1).
+    ylim : tuple of float, optional
+        Minimum and maximum values for the y-axis (Output 2).
+    r_scatter : float, default 1.0
+        The portion of data points displayed on the scatter plot (0 to 1).
 
     Returns
     -------
+    axs : Axes
+        Matplotlib axis for two-output graph.
     ax : Axes
         Matplotlib axis.
 
     """
     # needed to get the correct stacking order
     bins.columns = pd.RangeIndex(start=len(bins.columns), stop=0, step=-1)
 
+    if bins2 is not None:
+        fig, axs = plt.subplots(2, 2, sharex="col", sharey="row", figsize=(8, 8))
+        axs[0, 1].axis("off")
+
+        sns.histplot(
+            bins,
+            multiple="stack",
+            stat="probability",
+            palette=palette,
+            common_bins=True,
+            common_norm=True,
+            bins=n_bins,
+            legend=False,
+            ax=axs[0, 0],
+        )
+        axs[0, 0].set_xlim(xlim)
+        axs[0, 0].set_box_aspect(1)
+        axs[0, 0].axis("off")
+
+        data = pd.concat([pd.melt(bins), pd.melt(bins2)["value"]], axis=1)
+        data.columns = ["c", "x", "y"]
+
+        if r_scatter < 1.0:
+            data = data.sample(frac=r_scatter)
+
+        sns.scatterplot(
+            data=data,
+            x="x",
+            y="y",
+            hue="c",
+            palette=palette,
+            ax=axs[1, 0],
+            legend=False,
+        )
+        axs[1, 0].set(xlabel=output_name, ylabel=output_name2)
+        axs[1, 0].set_box_aspect(1)
+
+        sns.histplot(
+            data,
+            y="y",
+            hue="c",
+            multiple="stack",
+            stat="probability",
+            palette=palette,
+            common_bins=True,
+            common_norm=True,
+            bins=40,
+            legend=False,
+            ax=axs[1, 1],
+        )
+        axs[1, 1].set_ylim(ylim)
+        axs[1, 1].set_box_aspect(1)
+        axs[1, 1].axis("off")
+
+        fig.subplots_adjust(wspace=-0.015, hspace=0)
+        return axs[1, 0]
+
     if kind == "histogram":
         ax = sns.histplot(
             bins,

tests/test_visualization.py

@@ -0,0 +1,32 @@
+import pytest
+import pandas as pd
+import matplotlib.pyplot as plt
+import simdec as sd
+
+
+def test_visualization_single_output():
+    bins = pd.DataFrame({"s1": [1, 2], "s2": [3, 4]})
+    palette = [[1, 0, 0, 1], [0, 1, 0, 1]]
+
+    ax = sd.visualization(bins=bins, palette=palette, kind="histogram")
+    assert isinstance(ax, plt.Axes)
+
+    ax_box = sd.visualization(bins=bins, palette=palette, kind="boxplot")
+    assert isinstance(ax_box, plt.Axes)
+
+
+def test_visualization_two_outputs():
+    bins = pd.DataFrame({"s1": [1, 2]})
+    bins2 = pd.DataFrame({"s1": [5, 6]})
+    palette = [[1, 0, 0, 1]]
+
+    ax = sd.visualization(bins=bins, bins2=bins2, palette=palette)
+
+    assert ax.get_xlabel() == "Output 1"
+    assert len(ax.figure.axes) == 4
+
+
+def test_visualization_invalid_kind():
+    bins = pd.DataFrame({"s1": [1]})
+    with pytest.raises(ValueError, match="'kind' can only be 'histogram' or 'boxplot'"):
+        sd.visualization(bins=bins, palette=[[1, 0, 0, 1]], kind="invalid")