reporting, again aka python for people (was: History Quotes)

[Sébastien de Menten]

On Fri, Nov 20, 2015 at 10:58 PM, Wm... <tcnw81 at tarrcity.demon.co.uk> wrote:

> Wed, 18 Nov 2015 06:45:11 <CAB2pxDtda8MKvy98dGboq=FQ67PeTXZahWHDCtQ=
> CmxGHAMZfg at mail.gmail.com>
> Sébastien de Menten <sdementen at gmail.com> wrote...
>
> piecash 0.10 has been just released with the ability to get all
>>
> transactions/splits and prices into neat pandas DataFrames.
>
> I will try it this weekend with interest.
>
> nice, keep me informed !


> You can see an example of a usage of these dataframes to do basic reporting
> here :
> https://nbviewer.jupyter.org/github/sdementen/piecash/blob/master/example
> s/ipython/piecash_dataframes.ipynb
>
> That is looking wonderful, Sebastien, I am *very* impressed.  You've
> obviously put a lot of work in to only use the imports you have.
>
> Thanks Wm. The API is rather readable and simple to use. It can be a good
basis to generate reports (with pandas a.o.) or to generate documents (with
jinja2 and latex, or with reportlab a.o.).


> One Q while it is on my mind, why plotly rather than matplotlib ? <-- not
> a crit, a query, may become obvious when I try piecash latest myself
>
> Just because it has been open sourced last week and I wanted to try it :-)
But indeed matplotlib is probably a more mature solution for charts.
Anyway, there is no link between piecash and plotly (feel free to use
whatever you prefer), piecash just enables to export the splits and the
prices to pandas DataFrame. The chart were there to show visually the
results.


> I put the piecash google groups in CC so that further discussions specific
> to piecash can be held there instead of on the gnucash ML.
>
> I'm not sure they should really be separated, it is just the writing that
> is argued about, everything else is good for all, surely? :)



ok, I am fine with any preference from the ML