adding a diverging bar example to the horizontal bar documentation

doc/python/horizontal-bar-charts.md

@@ -217,6 +217,77 @@ fig.update_layout(annotations=annotations)
 fig.show()
 ```
 
+### Diverging Bar (or Butterfly) Chart
+
+Diverging bar charts show counts of positive outcomes or sentiments to the right of zero and counts of negative outcomes to the left of zero, allowing the reader to easily spot areas of excellence and concern.  Implementing presentation-ready versions of them in Plotly requires a few non standard layout and legendrank options.
+
+```
+import pandas as pd
+import plotly.graph_objects as go
+
+
+df = pd.read_csv('https://raw.githubusercontent.com/plotly/datasets/refs/heads/master/gss_2002_5_pt_likert.csv')
+#data source details are in this CSV file
+df.rename(columns={'Unnamed: 0':"Category"}, inplace=True)
+
+#achieve the diverging effect by putting a negative sign on the "disagree" answers 
+for v in ["Disagree","Strongly Disagree"]:
+    df[v]=df[v]*-1
+
+fig = go.Figure()
+# this color palette conveys meaning:  blues for negative, reds for positive, gray for Neither Agree nor Disagree
+color_by_category={
+    "Strongly Agree":'darkblue',
+    "Agree":'lightblue',
+    "Disagree":'orange',
+    "Strongly Disagree":'red',
+    "Neither Agree nor Disagree":'gray',
+}
+
+
+# We want the legend to be ordered in the same order that the categories appear, left to right --
+# which is different from the order in which we have to add the traces to the figure.
+# since we need to create the "somewhat" traces before the "strongly" traces to display
+# the segments in the desired order
+legend_rank_by_category={
+    "Strongly Disagree":1,
+    "Disagree":2,
+    "Agree":3,
+    "Strongly Agree":4,
+    "Neither Agree nor Disagree":5
+}
+# Add bars for each category
+for col in ["Disagree","Strongly Disagree","Agree","Strongly Agree"]:
+    fig.add_trace(go.Bar(
+        y=df["Category"], 
+        x=df[col], 
+        name=col, 
+        orientation='h',
+        marker=dict(color=color_by_category[col]),
+        legendrank=legend_rank_by_category[col]
+    ))
+
+fig.update_layout(
+   title="Reactions to statements from the 2002 General Social Survey:",
+    yaxis_title = "",
+    barmode='relative',  # Allows bars to diverge from the center
+    plot_bgcolor="white",
+)
+
+fig.update_xaxes(
+        title="Percent of Responses",
+        zeroline=True,  # Ensure there's a zero line for divergence
+        zerolinecolor="black",
+        # use array tick mode to show that the counts to the left of zero are still positive.
+        # this is hard coded; generalize this if you plan to create a function that takes unknown or widely varying data
+        tickmode = 'array',     
+        tickvals = [-50, 0, 50, 100],
+        ticktext = [50, 0, 50, 100]
+)
+
+fig.show()
+```
+
 ### Bar Chart with Line Plot
 
 ```python
@@ -335,4 +406,4 @@ fig.show()
 
 ### Reference
 
-See more examples of bar charts and styling options [here](https://plotly.com/python/bar-charts/).<br> See https://plotly.com/python/reference/bar/ for more information and chart attribute options!
+See more examples of bar charts and styling options [here](https://plotly.com/python/bar-charts/).<br> See https://plotly.com/python/reference/bar/ for more information and chart attribute options!

doc/python/performance.md

@@ -78,6 +78,10 @@ The following [array data types](https://numpy.org/devdocs/reference/arrays.scal
 
 Arrays or data types that are not supported for base64 encoding to Plotly.js's typed arrays specification will still work and render correctly with Plotly. Those arrays and or data types just won't have the performance benefits that Plotly.js's base64 typed arrays feature provides.
 
+### Dash Design Kit Compatibility
+
+The Dash Design Kit `Graph` component from versions of Dash Design Kit earlier than 1.13 doesn't work with Plotly.py version 6 when using Python objects listed in the previous "Arrays and Data Types Supported". Upgrade Dash Design Kit to 1.13 or later to resolve this issue.
+
 ### Unsupported Attributes
 
 Arrays passed to attributes with the following names are not supported for base64 encoding for rendering with Plotly.js.

doc/python/v6-changes.md

@@ -57,9 +57,12 @@ pip install anywidget
 
 Plotly.py now takes advantage of recent changes in how Plotly.js handles typed arrays for improved performance. See the [performance page](https://plotly.com/python/performance/) for more details.
 
+> If you are using Plotly.py 6 or later with Dash Design Kit, you may need to upgrade your Dash Design Kit version. See the [Dash Design Kit Compatibility section on the performance page](/python/performance/#dash-design-kit-compatibility) for more details.
+
+
 ## Dataframe Support
 
-Plotly Express now uses [Narwhals](https://narwhals-dev.github.io/narwhals/) to natively support pandas, Polars, and PyArrow. With this change, the [performance](https://plotly.com/python/performance/) of using Polars or PyArrow with Plotly Express is significantly improved. 
+Plotly Express now uses [Narwhals](https://narwhals-dev.github.io/narwhals/) to natively support pandas, Polars, and PyArrow. With this change, the [performance](https://plotly.com/python/performance/) of using Polars or PyArrow with Plotly Express is significantly improved.
 
 ## Mapbox Deprecation