Re: [Matplotlib-users] for Posterior predictive data distributions

[Paul H. <pmh...@gm...>]

(apologies if the list receives this twice)

On Fri, Jun 5, 2015 at 9:14 AM, Juan Wu <wuj...@gm...> wrote:
>
>> Hi, Experts,
>>
>> My colleagues and I have a question, how we can make a plot via python
>> like below. According to a guy's original paper, "Each panel shows the
>> normalized histograms of the observed data (bar plots) and the model
>> prediction (black lines) ".
>>
>> I believe that people can make it with Matplotlib. Any code suggestion
>> (with simple example data) would be much appreciated.
>>
>> (I am more comfortable with Matlab, but now the python code is preferred).
>>
>> J
>>
>

Juan,

It is, of course, very difficult to give any concrete advice without
knowing how your data are stored.

In any case, seaborn builds on matplotlib to provide some very advanced
visualization with a very concise API.

I recommend you look into the seaborn.distplot function and
seaborn.FacetGrid class.
http://web.stanford.edu/~mwaskom/software/seaborn/

-Paul