Halitim Bachir | 19 Jan 15:34 2016
Picon

Panda bar graph plot


Dear all,

I have some data taken at specific time of the day and i  want to display those data according to the time,
attached is the cvs file and the display output from pandas . I used the following script 

 v2=read_csv('v2_12.dat')  #data frame for v2
 v2.plot(kind='bar', x='Time_hhmmss', y='Av_phase',figsize=(12,1)) #display for v2 only

I want to see a gap on the display because there was no data recorded between  08:20:56  and   14:55:33      but on my display i see them side by  side 
Is there any way to do  this using panda display or in the pandas display option
Your help is highly appreciated 
Thanks much

FYI : my csv and output files are attached on this email


--
You received this message because you are subscribed to the Google Groups "PyData" group.
To unsubscribe from this group and stop receiving emails from it, send an email to pydata+unsubscribe-/JYPxA39Uh5TLH3MbocFFw@public.gmane.org.
For more options, visit https://groups.google.com/d/optout.
Attachment (v2_121.csv): text/csv, 4828 bytes
dartdog | 18 Jan 18:39 2016
Picon

A new add in for Dataframes provides some new formatting and totaling functionality (PrettyPandas)

FYI

https://github.com/HHammond/PrettyPandas

--
You received this message because you are subscribed to the Google Groups "PyData" group.
To unsubscribe from this group and stop receiving emails from it, send an email to pydata+unsubscribe-/JYPxA39Uh5TLH3MbocFFw@public.gmane.org.
For more options, visit https://groups.google.com/d/optout.
Yoel Shapiro | 17 Jan 08:40 2016
Picon

passing multiple DataFrames\Tables between pandas\matlab

hi

i'm looking for an elegant\lazy method to pass data from matlab to python and back to matlab

hdf (h5py) allows encapsulating the entire matlab workspace but i can't figure out how to read the Tables into DataFrames

saving tables to csv works well but my matlab-colleagues are using a struct that includes >10 tables and we don't want to start managing lists of files

to clarify, the matlab workspace looks something like this:

main_struct
- Table1
...
- Table10
- nested struct 1
- nested struct 2
- 4D array

--
You received this message because you are subscribed to the Google Groups "PyData" group.
To unsubscribe from this group and stop receiving emails from it, send an email to pydata+unsubscribe-/JYPxA39Uh5TLH3MbocFFw@public.gmane.org.
For more options, visit https://groups.google.com/d/optout.
Halitim Bachir | 17 Jan 15:46 2016
Picon

Hello Everybody I need to sort a dataframe according to a specific column the create new dataframes

Hello Everybody
I need to sort a dataframe according to a specific column then create new dataframes according to the sorted columns, each new created dataframe should contain an element from  list(set('the sorted header' )
any help please , i am new in python and pandas 
thanks much
Bachir

--
You received this message because you are subscribed to the Google Groups "PyData" group.
To unsubscribe from this group and stop receiving emails from it, send an email to pydata+unsubscribe-/JYPxA39Uh5TLH3MbocFFw@public.gmane.org.
For more options, visit https://groups.google.com/d/optout.
Wes McKinney | 16 Jan 18:01 2016
Picon
Gravatar

Governance for pandas development

hi everyone,

For those who aren't on the pandas-dev@... mailing list, I
wanted to make you aware that we've been discussing adopting and
publishing governance documents for pandas to bring it in line with
Jupyter and NumPy who have recently also done the same.

The first draft of these documents, which were adapted from Jupyter's
version (github.com/jupyter/governance), has been under discussion
here:

https://github.com/pydata/pandas-governance/pull/1

We also have been discussing adopting a code of conduct:

https://github.com/pydata/pandas-governance/pull/2

The hope with any governance documents is that it's a written version
of what is already happening in the development team, but it
additionally provides some clarity around things like the relationship
with NumFOCUS (which provides a conduit for financial support for
pandas).

The documents also describe how to become a part of pandas's core team
and what kind of behavior we expect from one another: e.g. openness
and transparency and minding potential conflicts of interest.

I welcome all to review the documents and raise comments or concerns
either on GitHub or here on the PyData mailing list.

all the best,
Wes

--

-- 
You received this message because you are subscribed to the Google Groups "PyData" group.
To unsubscribe from this group and stop receiving emails from it, send an email to pydata+unsubscribe@...
For more options, visit https://groups.google.com/d/optout.

'Michael' via PyData | 16 Jan 13:57 2016

status info in timeseries bar plot

In bar chart of a timeseries, status info shows:
x= , y=value

whatever y value is there

but why isn't x defined?
and how could we see the date in status bar?

like
x=date, y=value

is this issue pandas related? or matplotlib related?

--
You received this message because you are subscribed to the Google Groups "PyData" group.
To unsubscribe from this group and stop receiving emails from it, send an email to pydata+unsubscribe-/JYPxA39Uh5TLH3MbocFFw@public.gmane.org.
For more options, visit https://groups.google.com/d/optout.
questions anon | 12 Jan 00:23 2016
Picon

bar plot with rolling mean - issue with x-axis labelling

I am trying to combine a stacked bar plot with a rolling mean.
I have resolved issues in the past with the bar plot and x-axis labelling (thanks to this forum) but I seem to be having issues with my axis labelling again now that I am adding a rolling mean. 

I have attached two plots - one is just the stacked bar plot with the correct x-axis labelling and the other is when I include the rolling mean. It is like the rolling mean takes over the x-axis labelling?
Any feedback will be greatly appreciated.

result=pd.concat([df75, df50, df25],axis=1)

myplot=result.plot(kind='bar', stacked=True, color=['r','y','c'])

myplot.set_xticklabels([dt.strftime('%Y') for dt in result.index.to_pydatetime()])

plt.setp(myplot.get_xticklabels()[0::2], visible=False)

pd.rolling_mean(result, window=5).plot(ax=myplot,style='--', color=['r','y','c'])





--
You received this message because you are subscribed to the Google Groups "PyData" group.
To unsubscribe from this group and stop receiving emails from it, send an email to pydata+unsubscribe-/JYPxA39Uh5TLH3MbocFFw@public.gmane.org.
For more options, visit https://groups.google.com/d/optout.
ccxchen | 11 Jan 19:58 2016
Picon

how do a name the output data frame in pandas-ply

import re import os import csv import codecs import numpy as np import matplotlib as p import pdb import pandas as pd from pandas_ply import install_ply, X, sym_call install_ply(pd) (data_merged .groupby('index') .ply_select( count = X.index.count(), p_avg = X.item_price.mean() ))


Here
data_merged is the input dataframe but how do I name the output dataframe?

--
You received this message because you are subscribed to the Google Groups "PyData" group.
To unsubscribe from this group and stop receiving emails from it, send an email to pydata+unsubscribe-/JYPxA39Uh5TLH3MbocFFw@public.gmane.org.
For more options, visit https://groups.google.com/d/optout.
Oscar Martinez | 11 Jan 03:50 2016
Picon

Feedback Requested for Pandas-based Project

Hello all,

A while ago, I built a prototype of an application / dev tool that I continued to expand and use for practically every new project that I have worked on. While it might not be strictly about Pandas, it owes a whole lot to this project so I thought that I would seek some feedback here first. 

Every time worked on a new application that has some sort of client-server architecture and required data persistence (i.e. practically every web app out there), I found myself reinventing the wheel and doing a lot of boilerplate work just to get the interface between client and server running. Even though many of the tools for Python out there make it very easy to design routes for a RESTful API service, such as flask, django and bottle; persisting that information to disk and making that available to the client side of things becomes very repetitive across projects. During development, it makes little sense to spend a lot of time setting up and designing a proper database in addition to the API's, and in the early stages of a project things move so fast that the overhead does not justify the benefit.

Often times, I found myself using CSV files and handling a lot of the logic through Pandas and DataFrames because it was the path of least resistance. So I took it to the next logical step, and wrapped a prototyping database application around Pandas, which I creatively named PandasDatabase. Basically, it leverages Pandas to expose a very simple API that mimics the database-table-column model common to most database software; but keeps things simple and flexible enough to make early development painless. The idea is that, once there is a working prototype, the project can better design the API's and choose the production database software.

What do you think? Would you use it for prototyping?
Project URL: https://github.com/omtinez/pddb

Thank you,
Oscar

--
You received this message because you are subscribed to the Google Groups "PyData" group.
To unsubscribe from this group and stop receiving emails from it, send an email to pydata+unsubscribe-/JYPxA39Uh5TLH3MbocFFw@public.gmane.org.
For more options, visit https://groups.google.com/d/optout.
Paul Blelloch | 7 Jan 19:35 2016
Picon

.replace on a MultiIndex

This is probably trivial, but I'm not sure how to do it.  I have a MultiIndex, where I want to replace every occurrence of 't1' with 's' in level[1].  The .replace method only seems to work on the data itself and not the index, and while I could rebuild the index it seems that there should be simpler way.  Can I replace values in specific levels of a MultiIndex?

THANKS, Paul

--
You received this message because you are subscribed to the Google Groups "PyData" group.
To unsubscribe from this group and stop receiving emails from it, send an email to pydata+unsubscribe-/JYPxA39Uh5TLH3MbocFFw@public.gmane.org.
For more options, visit https://groups.google.com/d/optout.
Bryan Van de Ven | 6 Jan 22:48 2016
Gravatar

ANN: Bokeh 0.11 released

Hi all,

On behalf of the Bokeh team, I am excited to announce the release of version
0.11 of Bokeh.

Bokeh Version 0.11 is a large release with *many* new improvements. The
major focus of this release was to introduce a new Bokeh server, based on
Tornado and websockets that is more stable, has higher performance, and simpler
to use and deploy. You can already see a small sample of hosted Bokeh
application examples at a new site located here:

    http://demo.bokehplots.com

We will be adding many new examples here over the coming weeks.

There is so much exciting in this release, that we will need a few blog
posts to talk about everything. Keep an eye on our Twitter  <at> BokehPlots or the
mailing list for upcoming blog announcements. Some of the highlights from this
release are:

* New Bokeh Server based on Tornado and websockets
  - highly expanded documentation, with examples and guidance for usage
* User-Defined Models allowing anyone to extend Bokeh
* Significant GIS features and improvements
  - Support for Stamen, OpenStreetMap, and other tile sources
  - GeoJSON data source
  - Patches with holes
* WebGL support for rendering lines
* Python -> JS compilation for CustomJS callbacks (Py3 only for now)
* New general push_notebook() based on Jupyter comms
* Updates to charts, charts examples, and charts docs
* UX improvements
  - configurable and "auto" range bounds
  - wheel zoom scroll capture turned off by default
  - easily set visual styling for highlighting hovered points
  - responsive improvements, maintain plot aspect and auto-resize

Nearly 400 issues and PRs were closed for the release! See the CHANGELOG for full
details, and the release notes for known issues and any migration notes.

The next big feature we plan is to dramatically improve our layout options
using PhosporJS, which is also the foundation of the new Jupyter Workbench
project. We also plan to have several, more frequent point releases with
smaller incremental improvements and fixes.

I'd also like to take the opportunity to thank all the Bokeh contributors,
and especially new people who have helped greatly with this release: Havoc
Pennington, Greg Nordin, and Christian Tremblay.

If you are using Anaconda/miniconda, you can install it with conda:

    conda install bokeh

Alternatively, you can also install it with pip:

    pip install bokeh

Issues, enhancement requests, and pull requests can be made on the Bokeh
Github page: https://github.com/bokeh/bokeh

Full documentation is available at http://bokeh.pydata.org/en/0.11.0

Questions can be directed to the Bokeh mailing list: bokeh@...

Thanks,

The Bokeh Team

--

-- 
You received this message because you are subscribed to the Google Groups "PyData" group.
To unsubscribe from this group and stop receiving emails from it, send an email to pydata+unsubscribe@...
For more options, visit https://groups.google.com/d/optout.


Gmane