basic bokeh Bar chart

Question

In[21]: from bokeh.charts import Bar, output_file, show
In[22]: dict = {'sec': {u'A': 10, u'B': 20}}
In[23]: df = pd.DataFrame(dict)
In[24]: df
Out[24]: 
   sec
A   10
B   20
In[25]: output_file("bar.html")
In[26]: p = Bar(df)
In[27]: show(p)

I'd like a bar chart with A and B the labels, the 2 bars of size 10 & 20:

This example renders nothing.

I tried various variations with values=, labels= etc but it seems data is always aggregated by default (agg='sum')

Is Bar object too high level to plot what I'd like (2 bars, size 10 & 20 named A and B ) or am I using the object incorrectly ? Has the DataFrame I'm using the wrong 'format', should i put the index in a column ? if someone could provide an exemple of that very simple chart that would be awesome !

Edit: got it working with this, too bad index of the DataFrame cant be the labels, that seemed logical to me :)

import pandas as pd
from bokeh.charts import Bar
from bokeh.io import output_notebook, show
output_notebook()
dict = {'values': {u'A': 10, u'B': 20}}
df = pd.DataFrame(dict)
df['label'] = df.index
df
p = Bar(df, values='values',label='label')
show(p)

Answer 1

The other answers are outdated. The bokeh.charts API has been deprecated and removed, it should no longer be used. For basic (and not-so-basic) bar charts, users should now use the stable bokeh.plotting API as described in Handling Categorical Data. As an example:

from bokeh.io import show, output_file
from bokeh.plotting import figure

output_file("bars.html")

fruits = ['Apples', 'Pears', 'Nectarines', 'Plums', 'Grapes', 'Strawberries']

p = figure(x_range=fruits, plot_height=250, title="Fruit Counts")
p.vbar(x=fruits, top=[5, 3, 4, 2, 4, 6], width=0.9)

p.xgrid.grid_line_color = None
p.y_range.start = 0

show(p)

Which yields:

There are many more examples in the User's Guide link above showing how to achieve sorted, colormapped, stacked, grouped and nested bars, or work directly from Pandas DataFrames and GroupBys.