matplotlib user guide

In article <20130927140906.3ded39ae@...1166...>,

Many thanks for the feedback.

So ,my first cut was to segregate the user guide by topic. Below is
the summary I had in mind for an Introduction for New Users.

Hopefully this gives a flavour of what I have in mind.

I will attempt to put this into practice and post again when I have a
user guide coded that might work in my view.

mark

Introducing Plotting with Matplotlib

    Pyplot tutorial
        Controlling line properties
        Working with multiple figures and axes
        Working with text
    Interactive navigation
        Navigation Keyboard Shortcuts
    Working with text
        Text introduction
        Basic text commands
        Text properties and layout
        Writing mathematical expressions
        Text rendering With LaTeX
        Annotating text

...

On the whole this looks good to me. I so have a few comments, however:
- Would you be willing to include a section on using the class API? (I'm
assuming the above is all based on using pyplot?). I find there is a
huge gap between pyplot and the class API, and the documentation I've
found does little to bridge that gap.
- You have "Working with text" (including "annotating text") early on,
then "Legend guide" and "Annotating Axes" much later on. I wish these
were all together, as axes (values and labels), plot titles and legends
are arguably the main use cases for text in plots. Perhaps it would
suffice to have "Working with text" give a brief overview of how to do
each of these things, with pointers to the other sections for details.
An alternative is subsections within Working with text.
- In a similar vein: I'm surprised GridSpec comes before legends and
annotating axes.
- Please consider a section on 3-d plots.
- Please consider a section on animation.

-- Russell

···

mark <markh@...1166...> wrote:

+inf

Even better, dump pyplot and use primarily the OO API. Asside from
folks that dont want to change anything when moving from Matlab, we
should all be using teh primarily OO API.

is it really that hard to type:

ax.plot()

rather than

plot() ?

And when you move away from interactive use (and we all should fi your
typing more than 4-5 lines of code) teh OO interface is a much better
way to go.

(I know, iPython notebooks allow you do do a LOT with esentially an
interactive interpreter, but still.....)

Anyway, I've always thought it was a real shame that most of the
tutorials on MPL out there get people started on what I'm convinced is
the wrong foot.

- just my opinionated $0.02 worth...

-Chris

···

On Fri, Oct 4, 2013 at 1:40 PM, Russell E. Owen <rowen@...748...> wrote:

Introducing Plotting with Matplotlib

    Pyplot tutorial
        Controlling line properties
        Working with multiple figures and axes
        Working with text
    Interactive navigation
        Navigation Keyboard Shortcuts
    Working with text
        Text introduction
        Basic text commands
        Text properties and layout
        Writing mathematical expressions
        Text rendering With LaTeX
        Annotating text

...

- Would you be willing to include a section on using the class API? (I'm
assuming the above is all based on using pyplot?).

--

Christopher Barker, Ph.D.
Oceanographer

Emergency Response Division
NOAA/NOS/OR&R (206) 526-6959 voice
7600 Sand Point Way NE (206) 526-6329 fax
Seattle, WA 98115 (206) 526-6317 main reception

Chris.Barker@...236...

FWIW, I think my "Anatomy of Matplotlib" tutorial I gave at SciPy 2013
struck a balance between pyplot and the OO interface. I welcome any and all
feedback on that tutorial which I plan to give again next year as well as
an intermediate "Anatomy of Matplotlib" tutorial.

Cheers!
Ben Root

···

On Fri, Oct 18, 2013 at 2:13 PM, Chris Barker <chris.barker@...236...> wrote:

On Fri, Oct 4, 2013 at 1:40 PM, Russell E. Owen <rowen@...748...> wrote:
>> Introducing Plotting with Matplotlib
>>
>> Pyplot tutorial
>> Controlling line properties
>> Working with multiple figures and axes
>> Working with text
>> Interactive navigation
>> Navigation Keyboard Shortcuts
>> Working with text
>> Text introduction
>> Basic text commands
>> Text properties and layout
>> Writing mathematical expressions
>> Text rendering With LaTeX
>> Annotating text
> ...

> - Would you be willing to include a section on using the class API? (I'm
> assuming the above is all based on using pyplot?).

+inf

Even better, dump pyplot and use primarily the OO API. Asside from
folks that dont want to change anything when moving from Matlab, we
should all be using teh primarily OO API.

is it really that hard to type:

ax.plot()

rather than

plot() ?

And when you move away from interactive use (and we all should fi your
typing more than 4-5 lines of code) teh OO interface is a much better
way to go.

(I know, iPython notebooks allow you do do a LOT with esentially an
interactive interpreter, but still.....)

Anyway, I've always thought it was a real shame that most of the
tutorials on MPL out there get people started on what I'm convinced is
the wrong foot.

- just my opinionated $0.02 worth...

-Chris

Grat, I'll take a look.

Does the ipynb linked from the tutorial site have most of the
presentation material?

As it turns out, I need to give an intro to matplotlib class this week
-- I've been looking for some good materials to use -- why re-invent
the wheel?

I'll be sure to give you any feedback I have.

Hmmm.. this may be a time to put together a project of materials
designed to teach matplotlib in a classroom setting -- a little
different than a tutorial designed to be done on one's own.

There is a bunch of stuff scattered among scipy tutorials, bootcamp
lectures, etc, but having a central place to develop materials would
be nice.

-Chris

···

On Fri, Oct 18, 2013 at 11:56 AM, Benjamin Root <ben.root@...553...> wrote:

FWIW, I think my "Anatomy of Matplotlib" tutorial I gave at SciPy 2013
struck a balance between pyplot and the OO interface.

--

Christopher Barker, Ph.D.
Oceanographer

Emergency Response Division
NOAA/NOS/OR&R (206) 526-6959 voice
7600 Sand Point Way NE (206) 526-6329 fax
Seattle, WA 98115 (206) 526-6317 main reception

Chris.Barker@...236...

> FWIW, I think my "Anatomy of Matplotlib" tutorial I gave at SciPy 2013
> struck a balance between pyplot and the OO interface.

Grat, I'll take a look.

Does the ipynb linked from the tutorial site have most of the
presentation material?

Yup. Most of it is in the notebook. Here is the repo:
https://github.com/WeatherGod/AnatomyOfMatplotlib

As it turns out, I need to give an intro to matplotlib class this week
-- I've been looking for some good materials to use -- why re-invent
the wheel?

I'll be sure to give you any feedback I have.

Thanks! That would be appreciated! I am glad this will (hopefully) save
time and effort on your part to get others ramped up.

Hmmm.. this may be a time to put together a project of materials
designed to teach matplotlib in a classroom setting -- a little
different than a tutorial designed to be done on one's own.

There is a bunch of stuff scattered among scipy tutorials, bootcamp
lectures, etc, but having a central place to develop materials would
be nice.

-Chris

I agree with you in principle, but I think the question is "central to
whom?" The bootcamps are useful for their own purposes and audiences, while
the scipy tutorial s are useful for their own things, etc. At this point, I
think the bootcamps and SciPy and other conferences really should be the
best venues for pushing out the "classroom-oriented" type of tutorials.

Cheers!
Ben Root

···

On Mon, Oct 21, 2013 at 1:26 PM, Chris Barker <chris.barker@...236...> wrote:

On Fri, Oct 18, 2013 at 11:56 AM, Benjamin Root <ben.root@...553...> wrote: