proposed addition of a df.agg stacked go.bar example

doc/python/bar-charts.md

@@ -304,6 +304,56 @@ fig.update_layout(barmode='stack')
 fig.show()
 ```
 
+### Stacked Bar Chart From Aggregating a DataFrame 
+
+Stacked bar charts are a powerful way to present results summarizing categories generated using the Pandas aggregate commands. `pandas.DataFrame.agg` produces a wide data set format incompatible with `px.bar`. Transposing and updating the indexes to achieve `px.bar` compatibility is a somewhat involved option. Here is one straightforward alternative, which presents the aggregated data as a stacked bar using plotly.graph_objects.
+```
+
+from plotly import graph_objects as go
+import pandas as pd
+
+# Get one year of gapminder data
+url = 'https://raw.githubusercontent.com/plotly/datasets/master/gapminderDataFiveYear.csv'
+df = pd.read_csv(url)
+df = df[df['year']==2007]
+df["gdp"]=df["pop"]*df['gdpPercap']
+
+
+# Build the summary of interest
+df_summarized = df.groupby("continent", observed=True).agg("sum").reset_index()
+
+df_summarized["percent of world population"]=100*df_summarized["pop"]/df_summarized["pop"].sum()
+df_summarized["percent of world GDP"]=100*df_summarized["gdp"]/df_summarized["gdp"].sum()
+
+
+df = df_summarized[["continent", 
+"percent of world population",
+"percent of world GDP",
+]]
+
+# We now have a wide data frame, but it's in the opposite orientation from the one that px is designed to deal with.
+# Transposing it and rebuilding the indexes is an option, but iterating through the DF using graph objects is more succinct. 
+
+fig=go.Figure()
+for category in df_summarized["continent"].values:
+    fig.add_trace(go.Bar(
+            x=df.columns[1:],
+            # We need to get a pandas series that contains just the values to graph; 
+            # We do so by selecting the right row, selecting the right columns
+            # and then transposing and using iloc to convert to a series
+            # Here, we assume that the bar element category variable is in column 0
+            y=list(df.loc[df["continent"]==category][list(df.columns[1:])].transpose().iloc[:,0]),
+            name=str(category)
+
+
+        )
+)
+fig.update_layout(barmode="stack")
+
+fig.show()
+```
+
+
 ### Bar Chart with Hover Text
 
 ```python