I need to make Excel/Powerpoint-style charts.

[python]

Hi-


My office produces a lot of charts based on time-series data. We use MS
powerpoint and excel to make these charts. The data gets updated
regularly, but there doesn't seem to be an easy way to automate updating
the charts using these packages. So, we spend a non-trivial amount of
time manually updating lots of spreadsheets to redraw the charts we need.

I hope to find some sort of python graphics library that can make charts
similar to Excel and Powerpoint, and then I can start automating these
tasks.

Here is an example of the charts we make:

http://www.sarcastic-horse.com/example.pdf #52k size file.

Are there any packages that can do similar charts? I need to be able to
chart time-series data with occasional missing observations. I also need
to be able to shade some regions of the charts, and add text titles,
captions and notes.

I'm (slowly) learning the reportlab package, and it looks good, but I
wanted to check if there was an even better solution out there.

Thanks for the help.

 

[Eric Brunel]

Hi-


My office produces a lot of charts based on time-series data. We use MS
powerpoint and excel to make these charts. The data gets updated
regularly, but there doesn't seem to be an easy way to automate updating
the charts using these packages. So, we spend a non-trivial amount of
time manually updating lots of spreadsheets to redraw the charts we need.

I hope to find some sort of python graphics library that can make charts
similar to Excel and Powerpoint, and then I can start automating these
tasks.

Here is an example of the charts we make:

http://www.sarcastic-horse.com/example.pdf #52k size file.

Are there any packages that can do similar charts? I need to be able to
chart time-series data with occasional missing observations. I also need
to be able to shade some regions of the charts, and add text titles,
captions and notes.

I'm (slowly) learning the reportlab package, and it looks good, but I
wanted to check if there was an even better solution out there.

Thanks for the help.

Maybe you'll be interested in gdchart/PyGDChart: http://www.fred.net/brv/chart
I'm not really sure it can do everything you want, though...

HTH

 

[John J. Lee]

My office produces a lot of charts based on time-series data. We use MS
powerpoint and excel to make these charts. The data gets updated
regularly, but there doesn't seem to be an easy way to automate updating
the charts using these packages. So, we spend a non-trivial amount of
time manually updating lots of spreadsheets to redraw the charts we need.

Have you tried COM automation?

Get win32all, play with the example Excel COM automation code (there
may be some powerpoint code there too, dunno. Get confused, buy
Hammond & Robinson's book from O'Reilly (or download the free chapter
on COM from O'Reilly's site).

or...

I hope to find some sort of python graphics library that can make charts
similar to Excel and Powerpoint, and then I can start automating these
tasks.
http://www.reportlab.com/demos/pythonpoint/pythonpoint.html


Here is an example of the charts we make:

http://www.sarcastic-horse.com/example.pdf #52k size file.

Pythonpoint may be just the ticket -- it's built on top of Reportlab,
which is a nice pure-Python pdf generation library. pdf can do page
transitions and embed images, of course, and Reportlab & Pythonpoint
supports all that very conveniently.

Are there any packages that can do similar charts? I need to be able to
chart time-series data with occasional missing observations. I also need
to be able to shade some regions of the charts, and add text titles,
captions and notes.

The Pythonpoint page does mention charts, but I don't understand what
"custom diagrams and charts coded in various Python libraries" means
-- does Pythonpoint have chart support, or are they saying "this'll
work well with other Python charting libraries"?

My experience has been that open source scientific plotting libraries
are rather uneven in quality, but you may well have more luck with
more business-oriented stuff. Sticking with low-level reportlab API
might be simplest, of course.

I'm (slowly) learning the reportlab package, and it looks good, but I
wanted to check if there was an even better solution out there.

Oh -- so, did you already find Pythonpoint??


John

 

[R.Marquez]

Have you looked here?
http://www.python.org/topics/scicomp/plotting.html

-Ruben