matplotlib-devel Mailing List for matplotlib

I love this layout spec idea! Gridspec is a pain in the ass. Bonus points for actually drawing the letter e.g. on the top left corner of each panel, so that the figures are ready for publication.
On Thu, Mar 19, 2015 at 5:04 AM, Benjamin Root <ben...@ou...> wrote:
> I like that. Furthermore, now that we build and push the docs with every
> merge in master, there is less reason to not do it that way.
> On Wed, Mar 18, 2015 at 1:56 PM, Thomas Caswell <tca...@gm...> wrote:
>> Currently we are doing MEPs on the wiki (
>> https://github.com/matplotlib/matplotlib/wiki/MEPTemplate) , but I would
>> like to move them to be in the docs (make a MEP folder next to 'users'?) as
>> the visibility on the wiki is low, there isn't a great way to leave line
>> comments, and we should have these documents in the official docs
>> eventually.
>>
>> Tom
>>
>> On Wed, Mar 18, 2015 at 1:45 PM Nicolas P. Rougier <
>> Nic...@in...> wrote:
>>
>>>
>>> Yes, a MEP makes sense to discuss the proposal.
>>> What's the procedure to open a MEP (i.e. where) ?
>>>
>>> Nicolas
>>>
>>>
>>> > On 18 Mar 2015, at 18:44, Benjamin Root <ben...@ou...> wrote:
>>> >
>>> > Also, perhaps it makes sense to make this a MEP to finalize and
>>> document the spec?
>>> >
>>> > On Wed, Mar 18, 2015 at 1:42 PM, Benjamin Root <ben...@ou...> wrote:
>>> > That is neat. I would be sure to put in some "..seealso::" lines in
>>> places like plt.subplots and GridSpec and such.
>>> >
>>> > A thought... could this perhaps be extended somehow to specify
>>> colorbars in the layout? I am not sure how I would do that, but if we could
>>> come up with a way to do it, *that* would make this a killer feature. But
>>> even without that, this is still pretty useful.
>>> >
>>> > On Wed, Mar 18, 2015 at 1:20 PM, Thomas Caswell <tca...@gm...>
>>> wrote:
>>> > Cool. I think it make sense to put this in to `pyplot.py` next to
>>> `subplots`
>>> >
>>> > Tom
>>> >
>>> > On Wed, Mar 18, 2015 at 1:14 PM Nicolas P. Rougier <
>>> Nic...@in...> wrote:
>>> >
>>> > Hi,
>>> >
>>> > I've been experimenting with a simple idea for specifying plot layout
>>> in a rather intuitive way.
>>> > The idea is simply to "draw" your layout using strings.
>>> >
>>> > Examples:
>>> >
>>> > layout = ["AB"]
>>> > -> means two plots side by side with equal width
>>> >
>>> > layout = ["AAAB"]
>>> > -> means two plots side by side A being 3 times wider than B
>>> >
>>> > layout = ["AB",
>>> > "CC"]
>>> > -> means two plots (A & B) side by side and C below with full width
>>> >
>>> > layout = ["AB",
>>> > "C "]
>>> > -> means two plots (A & B) side by side and C below A (same width)
>>> >
>>> > etc... (have a look at sources)
>>> >
>>> > I guess you cannot express every layout but it might work for most
>>> common ones.
>>> >
>>> > If you think this might a good addition I can try to make a PR but I'm
>>> not sure where to insert it.
>>> > My idea would be to have a layout function such that you can write:
>>> >
>>> > A,B,C = plt.layout(["AB", "CC"], border=0.01)
>>> > A.plot(...)
>>> > B.plot(...)
>>> > C.plot(...)
>>> >
>>> >
>>> > Nicolas
>>> >
>>> >
>>> > ------------------------------------------------------------
>>> ------------------
>>> > Dive into the World of Parallel Programming The Go Parallel Website,
>>> sponsored
>>> > by Intel and developed in partnership with Slashdot Media, is your hub
>>> for all
>>> > things parallel software development, from weekly thought leadership
>>> blogs to
>>> > news, videos, case studies, tutorials and more. Take a look and join the
>>> > conversation now. http://goparallel.sourceforge.
>>> net/_______________________________________________
>>> > Matplotlib-devel mailing list
>>> > Mat...@li...
>>> > https://lists.sourceforge.net/lists/listinfo/matplotlib-devel
>>> >
>>> > ------------------------------------------------------------
>>> ------------------
>>> > Dive into the World of Parallel Programming The Go Parallel Website,
>>> sponsored
>>> > by Intel and developed in partnership with Slashdot Media, is your hub
>>> for all
>>> > things parallel software development, from weekly thought leadership
>>> blogs to
>>> > news, videos, case studies, tutorials and more. Take a look and join the
>>> > conversation now. http://goparallel.sourceforge.net/
>>> > _______________________________________________
>>> > Matplotlib-devel mailing list
>>> > Mat...@li...
>>> > https://lists.sourceforge.net/lists/listinfo/matplotlib-devel
>>> >
>>> >
>>> >
>>>
>>>

I like that. Furthermore, now that we build and push the docs with every
merge in master, there is less reason to not do it that way.
On Wed, Mar 18, 2015 at 1:56 PM, Thomas Caswell <tca...@gm...> wrote:
> Currently we are doing MEPs on the wiki (
> https://github.com/matplotlib/matplotlib/wiki/MEPTemplate) , but I would
> like to move them to be in the docs (make a MEP folder next to 'users'?) as
> the visibility on the wiki is low, there isn't a great way to leave line
> comments, and we should have these documents in the official docs
> eventually.
>
> Tom
>
> On Wed, Mar 18, 2015 at 1:45 PM Nicolas P. Rougier <
> Nic...@in...> wrote:
>
>>
>> Yes, a MEP makes sense to discuss the proposal.
>> What's the procedure to open a MEP (i.e. where) ?
>>
>> Nicolas
>>
>>
>> > On 18 Mar 2015, at 18:44, Benjamin Root <ben...@ou...> wrote:
>> >
>> > Also, perhaps it makes sense to make this a MEP to finalize and
>> document the spec?
>> >
>> > On Wed, Mar 18, 2015 at 1:42 PM, Benjamin Root <ben...@ou...> wrote:
>> > That is neat. I would be sure to put in some "..seealso::" lines in
>> places like plt.subplots and GridSpec and such.
>> >
>> > A thought... could this perhaps be extended somehow to specify
>> colorbars in the layout? I am not sure how I would do that, but if we could
>> come up with a way to do it, *that* would make this a killer feature. But
>> even without that, this is still pretty useful.
>> >
>> > On Wed, Mar 18, 2015 at 1:20 PM, Thomas Caswell <tca...@gm...>
>> wrote:
>> > Cool. I think it make sense to put this in to `pyplot.py` next to
>> `subplots`
>> >
>> > Tom
>> >
>> > On Wed, Mar 18, 2015 at 1:14 PM Nicolas P. Rougier <
>> Nic...@in...> wrote:
>> >
>> > Hi,
>> >
>> > I've been experimenting with a simple idea for specifying plot layout
>> in a rather intuitive way.
>> > The idea is simply to "draw" your layout using strings.
>> >
>> > Examples:
>> >
>> > layout = ["AB"]
>> > -> means two plots side by side with equal width
>> >
>> > layout = ["AAAB"]
>> > -> means two plots side by side A being 3 times wider than B
>> >
>> > layout = ["AB",
>> > "CC"]
>> > -> means two plots (A & B) side by side and C below with full width
>> >
>> > layout = ["AB",
>> > "C "]
>> > -> means two plots (A & B) side by side and C below A (same width)
>> >
>> > etc... (have a look at sources)
>> >
>> > I guess you cannot express every layout but it might work for most
>> common ones.
>> >
>> > If you think this might a good addition I can try to make a PR but I'm
>> not sure where to insert it.
>> > My idea would be to have a layout function such that you can write:
>> >
>> > A,B,C = plt.layout(["AB", "CC"], border=0.01)
>> > A.plot(...)
>> > B.plot(...)
>> > C.plot(...)
>> >
>> >
>> > Nicolas
>> >
>> >
>> > ------------------------------------------------------------
>> ------------------
>> > Dive into the World of Parallel Programming The Go Parallel Website,
>> sponsored
>> > by Intel and developed in partnership with Slashdot Media, is your hub
>> for all
>> > things parallel software development, from weekly thought leadership
>> blogs to
>> > news, videos, case studies, tutorials and more. Take a look and join the
>> > conversation now. http://goparallel.sourceforge.
>> net/_______________________________________________
>> > Matplotlib-devel mailing list
>> > Mat...@li...
>> > https://lists.sourceforge.net/lists/listinfo/matplotlib-devel
>> >
>> > ------------------------------------------------------------
>> ------------------
>> > Dive into the World of Parallel Programming The Go Parallel Website,
>> sponsored
>> > by Intel and developed in partnership with Slashdot Media, is your hub
>> for all
>> > things parallel software development, from weekly thought leadership
>> blogs to
>> > news, videos, case studies, tutorials and more. Take a look and join the
>> > conversation now. http://goparallel.sourceforge.net/
>> > _______________________________________________
>> > Matplotlib-devel mailing list
>> > Mat...@li...
>> > https://lists.sourceforge.net/lists/listinfo/matplotlib-devel
>> >
>> >
>> >
>>
>>

Currently we are doing MEPs on the wiki (
https://github.com/matplotlib/matplotlib/wiki/MEPTemplate) , but I would
like to move them to be in the docs (make a MEP folder next to 'users'?) as
the visibility on the wiki is low, there isn't a great way to leave line
comments, and we should have these documents in the official docs
eventually.
Tom
On Wed, Mar 18, 2015 at 1:45 PM Nicolas P. Rougier <Nic...@in...>
wrote:
>
> Yes, a MEP makes sense to discuss the proposal.
> What's the procedure to open a MEP (i.e. where) ?
>
> Nicolas
>
>
> > On 18 Mar 2015, at 18:44, Benjamin Root <ben...@ou...> wrote:
> >
> > Also, perhaps it makes sense to make this a MEP to finalize and document
> the spec?
> >
> > On Wed, Mar 18, 2015 at 1:42 PM, Benjamin Root <ben...@ou...> wrote:
> > That is neat. I would be sure to put in some "..seealso::" lines in
> places like plt.subplots and GridSpec and such.
> >
> > A thought... could this perhaps be extended somehow to specify colorbars
> in the layout? I am not sure how I would do that, but if we could come up
> with a way to do it, *that* would make this a killer feature. But even
> without that, this is still pretty useful.
> >
> > On Wed, Mar 18, 2015 at 1:20 PM, Thomas Caswell <tca...@gm...>
> wrote:
> > Cool. I think it make sense to put this in to `pyplot.py` next to
> `subplots`
> >
> > Tom
> >
> > On Wed, Mar 18, 2015 at 1:14 PM Nicolas P. Rougier <
> Nic...@in...> wrote:
> >
> > Hi,
> >
> > I've been experimenting with a simple idea for specifying plot layout in
> a rather intuitive way.
> > The idea is simply to "draw" your layout using strings.
> >
> > Examples:
> >
> > layout = ["AB"]
> > -> means two plots side by side with equal width
> >
> > layout = ["AAAB"]
> > -> means two plots side by side A being 3 times wider than B
> >
> > layout = ["AB",
> > "CC"]
> > -> means two plots (A & B) side by side and C below with full width
> >
> > layout = ["AB",
> > "C "]
> > -> means two plots (A & B) side by side and C below A (same width)
> >
> > etc... (have a look at sources)
> >
> > I guess you cannot express every layout but it might work for most
> common ones.
> >
> > If you think this might a good addition I can try to make a PR but I'm
> not sure where to insert it.
> > My idea would be to have a layout function such that you can write:
> >
> > A,B,C = plt.layout(["AB", "CC"], border=0.01)
> > A.plot(...)
> > B.plot(...)
> > C.plot(...)
> >
> >
> > Nicolas
> >
> >
> > ------------------------------------------------------------
> ------------------
> > Dive into the World of Parallel Programming The Go Parallel Website,
> sponsored
> > by Intel and developed in partnership with Slashdot Media, is your hub
> for all
> > things parallel software development, from weekly thought leadership
> blogs to
> > news, videos, case studies, tutorials and more. Take a look and join the
> > conversation now. http://goparallel.sourceforge.
> net/_______________________________________________
> > Matplotlib-devel mailing list
> > Mat...@li...
> > https://lists.sourceforge.net/lists/listinfo/matplotlib-devel
> >
> > ------------------------------------------------------------
> ------------------
> > Dive into the World of Parallel Programming The Go Parallel Website,
> sponsored
> > by Intel and developed in partnership with Slashdot Media, is your hub
> for all
> > things parallel software development, from weekly thought leadership
> blogs to
> > news, videos, case studies, tutorials and more. Take a look and join the
> > conversation now. http://goparallel.sourceforge.net/
> > _______________________________________________
> > Matplotlib-devel mailing list
> > Mat...@li...
> > https://lists.sourceforge.net/lists/listinfo/matplotlib-devel
> >
> >
> >
>
>

Fantastic question. I don't see any mention of it in the docs. Perhaps I
should create a MEP to discuss that...
This is the best I could find: https://github.com/matplotlib/matplotlib/wiki
On Wed, Mar 18, 2015 at 1:45 PM, Nicolas P. Rougier <
Nic...@in...> wrote:
>
> Yes, a MEP makes sense to discuss the proposal.
> What's the procedure to open a MEP (i.e. where) ?
>
> Nicolas
>
>
> > On 18 Mar 2015, at 18:44, Benjamin Root <ben...@ou...> wrote:
> >
> > Also, perhaps it makes sense to make this a MEP to finalize and document
> the spec?
> >
> > On Wed, Mar 18, 2015 at 1:42 PM, Benjamin Root <ben...@ou...> wrote:
> > That is neat. I would be sure to put in some "..seealso::" lines in
> places like plt.subplots and GridSpec and such.
> >
> > A thought... could this perhaps be extended somehow to specify colorbars
> in the layout? I am not sure how I would do that, but if we could come up
> with a way to do it, *that* would make this a killer feature. But even
> without that, this is still pretty useful.
> >
> > On Wed, Mar 18, 2015 at 1:20 PM, Thomas Caswell <tca...@gm...>
> wrote:
> > Cool. I think it make sense to put this in to `pyplot.py` next to
> `subplots`
> >
> > Tom
> >
> > On Wed, Mar 18, 2015 at 1:14 PM Nicolas P. Rougier <
> Nic...@in...> wrote:
> >
> > Hi,
> >
> > I've been experimenting with a simple idea for specifying plot layout in
> a rather intuitive way.
> > The idea is simply to "draw" your layout using strings.
> >
> > Examples:
> >
> > layout = ["AB"]
> > -> means two plots side by side with equal width
> >
> > layout = ["AAAB"]
> > -> means two plots side by side A being 3 times wider than B
> >
> > layout = ["AB",
> > "CC"]
> > -> means two plots (A & B) side by side and C below with full width
> >
> > layout = ["AB",
> > "C "]
> > -> means two plots (A & B) side by side and C below A (same width)
> >
> > etc... (have a look at sources)
> >
> > I guess you cannot express every layout but it might work for most
> common ones.
> >
> > If you think this might a good addition I can try to make a PR but I'm
> not sure where to insert it.
> > My idea would be to have a layout function such that you can write:
> >
> > A,B,C = plt.layout(["AB", "CC"], border=0.01)
> > A.plot(...)
> > B.plot(...)
> > C.plot(...)
> >
> >
> > Nicolas
> >
> >
> >
> ------------------------------------------------------------------------------
> > Dive into the World of Parallel Programming The Go Parallel Website,
> sponsored
> > by Intel and developed in partnership with Slashdot Media, is your hub
> for all
> > things parallel software development, from weekly thought leadership
> blogs to
> > news, videos, case studies, tutorials and more. Take a look and join the
> > conversation now.
> http://goparallel.sourceforge.net/_______________________________________________
> > Matplotlib-devel mailing list
> > Mat...@li...
> > https://lists.sourceforge.net/lists/listinfo/matplotlib-devel
> >
> >
> ------------------------------------------------------------------------------
> > Dive into the World of Parallel Programming The Go Parallel Website,
> sponsored
> > by Intel and developed in partnership with Slashdot Media, is your hub
> for all
> > things parallel software development, from weekly thought leadership
> blogs to
> > news, videos, case studies, tutorials and more. Take a look and join the
> > conversation now. http://goparallel.sourceforge.net/
> > _______________________________________________
> > Matplotlib-devel mailing list
> > Mat...@li...
> > https://lists.sourceforge.net/lists/listinfo/matplotlib-devel
> >
> >
> >
>
>
>
> ------------------------------------------------------------------------------
> Dive into the World of Parallel Programming The Go Parallel Website,
> sponsored
> by Intel and developed in partnership with Slashdot Media, is your hub for
> all
> things parallel software development, from weekly thought leadership blogs
> to
> news, videos, case studies, tutorials and more. Take a look and join the
> conversation now. http://goparallel.sourceforge.net/
> _______________________________________________
> Matplotlib-devel mailing list
> Mat...@li...
> https://lists.sourceforge.net/lists/listinfo/matplotlib-devel
>

On 2015年03月18日 7:42 AM, Benjamin Root wrote:
> A thought... could this perhaps be extended somehow to specify colorbars
> in the layout?
A lower-case letter could indicate a colorbar-size Axes:
layout = ["ABc",
 "DE ",
 "ff "]
would put a vertical think axes to the right of B, and a double-wide 
hoizontal one below D and E.
All of this seems like an alternative API for gridspec and axes_grid1.
I am concerned about ending up with too many ways to do things, but with 
subtle differences.
How much control over spacing and sizing would be provided by kwargs or 
other adjustment mechanisms? How would this relate to subplot_params?
Eric

Yes, a MEP makes sense to discuss the proposal.
What's the procedure to open a MEP (i.e. where) ?
Nicolas
> On 18 Mar 2015, at 18:44, Benjamin Root <ben...@ou...> wrote:
> 
> Also, perhaps it makes sense to make this a MEP to finalize and document the spec?
> 
> On Wed, Mar 18, 2015 at 1:42 PM, Benjamin Root <ben...@ou...> wrote:
> That is neat. I would be sure to put in some "..seealso::" lines in places like plt.subplots and GridSpec and such.
> 
> A thought... could this perhaps be extended somehow to specify colorbars in the layout? I am not sure how I would do that, but if we could come up with a way to do it, *that* would make this a killer feature. But even without that, this is still pretty useful.
> 
> On Wed, Mar 18, 2015 at 1:20 PM, Thomas Caswell <tca...@gm...> wrote:
> Cool. I think it make sense to put this in to `pyplot.py` next to `subplots`
> 
> Tom
> 
> On Wed, Mar 18, 2015 at 1:14 PM Nicolas P. Rougier <Nic...@in...> wrote:
> 
> Hi,
> 
> I've been experimenting with a simple idea for specifying plot layout in a rather intuitive way.
> The idea is simply to "draw" your layout using strings.
> 
> Examples:
> 
> layout = ["AB"]
> -> means two plots side by side with equal width
> 
> layout = ["AAAB"]
> -> means two plots side by side A being 3 times wider than B
> 
> layout = ["AB",
> "CC"]
> -> means two plots (A & B) side by side and C below with full width
> 
> layout = ["AB",
> "C "]
> -> means two plots (A & B) side by side and C below A (same width)
> 
> etc... (have a look at sources)
> 
> I guess you cannot express every layout but it might work for most common ones.
> 
> If you think this might a good addition I can try to make a PR but I'm not sure where to insert it.
> My idea would be to have a layout function such that you can write:
> 
> A,B,C = plt.layout(["AB", "CC"], border=0.01)
> A.plot(...)
> B.plot(...)
> C.plot(...)
> 
> 
> Nicolas
> 
> 
> ------------------------------------------------------------------------------
> Dive into the World of Parallel Programming The Go Parallel Website, sponsored
> by Intel and developed in partnership with Slashdot Media, is your hub for all
> things parallel software development, from weekly thought leadership blogs to
> news, videos, case studies, tutorials and more. Take a look and join the
> conversation now. http://goparallel.sourceforge.net/_______________________________________________
> Matplotlib-devel mailing list
> Mat...@li...
> https://lists.sourceforge.net/lists/listinfo/matplotlib-devel
> 
> ------------------------------------------------------------------------------
> Dive into the World of Parallel Programming The Go Parallel Website, sponsored
> by Intel and developed in partnership with Slashdot Media, is your hub for all
> things parallel software development, from weekly thought leadership blogs to
> news, videos, case studies, tutorials and more. Take a look and join the
> conversation now. http://goparallel.sourceforge.net/
> _______________________________________________
> Matplotlib-devel mailing list
> Mat...@li...
> https://lists.sourceforge.net/lists/listinfo/matplotlib-devel
> 
> 
> 

Also, perhaps it makes sense to make this a MEP to finalize and document
the spec?
On Wed, Mar 18, 2015 at 1:42 PM, Benjamin Root <ben...@ou...> wrote:
> That is neat. I would be sure to put in some "..seealso::" lines in places
> like plt.subplots and GridSpec and such.
>
> A thought... could this perhaps be extended somehow to specify colorbars
> in the layout? I am not sure how I would do that, but if we could come up
> with a way to do it, *that* would make this a killer feature. But even
> without that, this is still pretty useful.
>
> On Wed, Mar 18, 2015 at 1:20 PM, Thomas Caswell <tca...@gm...>
> wrote:
>
>> Cool. I think it make sense to put this in to `pyplot.py` next to
>> `subplots`
>>
>> Tom
>>
>> On Wed, Mar 18, 2015 at 1:14 PM Nicolas P. Rougier <
>> Nic...@in...> wrote:
>>
>>>
>>> Hi,
>>>
>>> I've been experimenting with a simple idea for specifying plot layout in
>>> a rather intuitive way.
>>> The idea is simply to "draw" your layout using strings.
>>>
>>> Examples:
>>>
>>> layout = ["AB"]
>>> -> means two plots side by side with equal width
>>>
>>> layout = ["AAAB"]
>>> -> means two plots side by side A being 3 times wider than B
>>>
>>> layout = ["AB",
>>> "CC"]
>>> -> means two plots (A & B) side by side and C below with full width
>>>
>>> layout = ["AB",
>>> "C "]
>>> -> means two plots (A & B) side by side and C below A (same width)
>>>
>>> etc... (have a look at sources)
>>>
>>> I guess you cannot express every layout but it might work for most
>>> common ones.
>>>
>>> If you think this might a good addition I can try to make a PR but I'm
>>> not sure where to insert it.
>>> My idea would be to have a layout function such that you can write:
>>>
>>> A,B,C = plt.layout(["AB", "CC"], border=0.01)
>>> A.plot(...)
>>> B.plot(...)
>>> C.plot(...)
>>>
>>>
>>> Nicolas
>>>
>>>
>>> ------------------------------------------------------------
>>> ------------------
>>> Dive into the World of Parallel Programming The Go Parallel Website,
>>> sponsored
>>> by Intel and developed in partnership with Slashdot Media, is your hub
>>> for all
>>> things parallel software development, from weekly thought leadership
>>> blogs to
>>> news, videos, case studies, tutorials and more. Take a look and join the
>>> conversation now. http://goparallel.sourceforge.net/
>>> _______________________________________________
>>> Matplotlib-devel mailing list
>>> Mat...@li...
>>> https://lists.sourceforge.net/lists/listinfo/matplotlib-devel
>>>
>>
>>
>> ------------------------------------------------------------------------------
>> Dive into the World of Parallel Programming The Go Parallel Website,
>> sponsored
>> by Intel and developed in partnership with Slashdot Media, is your hub
>> for all
>> things parallel software development, from weekly thought leadership
>> blogs to
>> news, videos, case studies, tutorials and more. Take a look and join the
>> conversation now. http://goparallel.sourceforge.net/
>> _______________________________________________
>> Matplotlib-devel mailing list
>> Mat...@li...
>> https://lists.sourceforge.net/lists/listinfo/matplotlib-devel
>>
>>
>

That is neat. I would be sure to put in some "..seealso::" lines in places
like plt.subplots and GridSpec and such.
A thought... could this perhaps be extended somehow to specify colorbars in
the layout? I am not sure how I would do that, but if we could come up with
a way to do it, *that* would make this a killer feature. But even without
that, this is still pretty useful.
On Wed, Mar 18, 2015 at 1:20 PM, Thomas Caswell <tca...@gm...> wrote:
> Cool. I think it make sense to put this in to `pyplot.py` next to
> `subplots`
>
> Tom
>
> On Wed, Mar 18, 2015 at 1:14 PM Nicolas P. Rougier <
> Nic...@in...> wrote:
>
>>
>> Hi,
>>
>> I've been experimenting with a simple idea for specifying plot layout in
>> a rather intuitive way.
>> The idea is simply to "draw" your layout using strings.
>>
>> Examples:
>>
>> layout = ["AB"]
>> -> means two plots side by side with equal width
>>
>> layout = ["AAAB"]
>> -> means two plots side by side A being 3 times wider than B
>>
>> layout = ["AB",
>> "CC"]
>> -> means two plots (A & B) side by side and C below with full width
>>
>> layout = ["AB",
>> "C "]
>> -> means two plots (A & B) side by side and C below A (same width)
>>
>> etc... (have a look at sources)
>>
>> I guess you cannot express every layout but it might work for most common
>> ones.
>>
>> If you think this might a good addition I can try to make a PR but I'm
>> not sure where to insert it.
>> My idea would be to have a layout function such that you can write:
>>
>> A,B,C = plt.layout(["AB", "CC"], border=0.01)
>> A.plot(...)
>> B.plot(...)
>> C.plot(...)
>>
>>
>> Nicolas
>>
>>
>> ------------------------------------------------------------
>> ------------------
>> Dive into the World of Parallel Programming The Go Parallel Website,
>> sponsored
>> by Intel and developed in partnership with Slashdot Media, is your hub
>> for all
>> things parallel software development, from weekly thought leadership
>> blogs to
>> news, videos, case studies, tutorials and more. Take a look and join the
>> conversation now. http://goparallel.sourceforge.net/
>> _______________________________________________
>> Matplotlib-devel mailing list
>> Mat...@li...
>> https://lists.sourceforge.net/lists/listinfo/matplotlib-devel
>>
>
>
> ------------------------------------------------------------------------------
> Dive into the World of Parallel Programming The Go Parallel Website,
> sponsored
> by Intel and developed in partnership with Slashdot Media, is your hub for
> all
> things parallel software development, from weekly thought leadership blogs
> to
> news, videos, case studies, tutorials and more. Take a look and join the
> conversation now. http://goparallel.sourceforge.net/
> _______________________________________________
> Matplotlib-devel mailing list
> Mat...@li...
> https://lists.sourceforge.net/lists/listinfo/matplotlib-devel
>
>

Cool. I think it make sense to put this in to `pyplot.py` next to
`subplots`
Tom
On Wed, Mar 18, 2015 at 1:14 PM Nicolas P. Rougier <Nic...@in...>
wrote:
>
> Hi,
>
> I've been experimenting with a simple idea for specifying plot layout in a
> rather intuitive way.
> The idea is simply to "draw" your layout using strings.
>
> Examples:
>
> layout = ["AB"]
> -> means two plots side by side with equal width
>
> layout = ["AAAB"]
> -> means two plots side by side A being 3 times wider than B
>
> layout = ["AB",
> "CC"]
> -> means two plots (A & B) side by side and C below with full width
>
> layout = ["AB",
> "C "]
> -> means two plots (A & B) side by side and C below A (same width)
>
> etc... (have a look at sources)
>
> I guess you cannot express every layout but it might work for most common
> ones.
>
> If you think this might a good addition I can try to make a PR but I'm not
> sure where to insert it.
> My idea would be to have a layout function such that you can write:
>
> A,B,C = plt.layout(["AB", "CC"], border=0.01)
> A.plot(...)
> B.plot(...)
> C.plot(...)
>
>
> Nicolas
>
>
> ------------------------------------------------------------
> ------------------
> Dive into the World of Parallel Programming The Go Parallel Website,
> sponsored
> by Intel and developed in partnership with Slashdot Media, is your hub for
> all
> things parallel software development, from weekly thought leadership blogs
> to
> news, videos, case studies, tutorials and more. Take a look and join the
> conversation now. http://goparallel.sourceforge.net/
> _______________________________________________
> Matplotlib-devel mailing list
> Mat...@li...
> https://lists.sourceforge.net/lists/listinfo/matplotlib-devel
>

Hi,
I've been experimenting with a simple idea for specifying plot layout in a rather intuitive way.
The idea is simply to "draw" your layout using strings.
Examples:
layout = ["AB"]
-> means two plots side by side with equal width
layout = ["AAAB"]
-> means two plots side by side A being 3 times wider than B
layout = ["AB",
 "CC"]
-> means two plots (A & B) side by side and C below with full width
layout = ["AB",
 "C "]
-> means two plots (A & B) side by side and C below A (same width)
etc... (have a look at sources)
I guess you cannot express every layout but it might work for most common ones.
If you think this might a good addition I can try to make a PR but I'm not sure where to insert it.
My idea would be to have a layout function such that you can write:
A,B,C = plt.layout(["AB", "CC"], border=0.01)
A.plot(...)
B.plot(...)
C.plot(...)
Nicolas

Keith,
Back to the issue at hand. could you do the following?
import setuptools
print(setuptools.__file__)
It would be interesting to see if that path differs from the path of the
egg you just listed.
Ben
On Wed, Mar 18, 2015 at 11:39 AM, <kei...@bt...> wrote:
> pip still thinks I have the latest. I think it's a question of how to
> force the matplotlib setup.py to use actually it.
> Keith
>
> kbriggs:~/Downloads/matplotlib-1.4.3> sudo pip3 install setuptools
> --upgrade
> Requirement already up-to-date: setuptools in
> /usr/local/lib/python3.4/dist-packages/setuptools-14.3-py3.4.egg
> Cleaning up...
>
> ________________________________________
> From: ben...@gm... [ben...@gm...] On Behalf Of Benjamin
> Root [ben...@ou...]
> Sent: 18 March 2015 15:33
> To: Briggs,KM,Keith,TUB2 R
> Cc: matplotlib development list
> Subject: Re: [matplotlib-devel] 1.4.3 does not build on Ubuntu 14 with
> python3
>
> I would just use pip. Ubuntu/Debian has really messed up the python
> environment in more ways than one.
>
> On Wed, Mar 18, 2015 at 11:08 AM, keithbriggs <kei...@bt...<mailto:
> kei...@bt...>> wrote:
> The Ubuntu package manager tells me it is up to date.
> If I download setuptools-14.3 and install, it goes into /usr/local/lib/ and
> doesn't get used.
> How do I force it to be used?
> Or am I supposed to override the Ubuntu package manager?
> Keith
>
>
>
>
>
>
> ------------------------------------------------------------------------------
> Dive into the World of Parallel Programming The Go Parallel Website,
> sponsored
> by Intel and developed in partnership with Slashdot Media, is your hub for
> all
> things parallel software development, from weekly thought leadership blogs
> to
> news, videos, case studies, tutorials and more. Take a look and join the
> conversation now. http://goparallel.sourceforge.net/
> _______________________________________________
> Matplotlib-devel mailing list
> Mat...@li...
> https://lists.sourceforge.net/lists/listinfo/matplotlib-devel
>

On Wed, Mar 18, 2015 at 3:51 PM, Benjamin Root <ben...@ou...> wrote:
> Let's not turn this into a flame war and let's figure out why this user's
> setuptools is not being recognized properly. We can discuss my experience
> with python2.7 on 12.04 LTS off-list, if you wish (i personally switched to
> miniconda to avoid the hassle).
I agree, but you started it :) also if you had problems with Ubuntu, I
would have avoided to bring Debian in the discussion - case closed for
me.
Cheers,
-- 
Sandro Tosi (aka morph, morpheus, matrixhasu)
My website: http://matrixhasu.altervista.org/
Me at Debian: http://wiki.debian.org/SandroTosi

Let's not turn this into a flame war and let's figure out why this user's
setuptools is not being recognized properly. We can discuss my experience
with python2.7 on 12.04 LTS off-list, if you wish (i personally switched to
miniconda to avoid the hassle).
On Wed, Mar 18, 2015 at 11:43 AM, Sandro Tosi <mo...@de...> wrote:
> On Wed, Mar 18, 2015 at 3:33 PM, Benjamin Root <ben...@ou...> wrote:
> > I would just use pip. Ubuntu/Debian has really messed up the python
> > environment in more ways than one.
>
> <debian developer hat=on/> ehm... either you back up your claims
> (and/or try to understand the reasons behind the distro needs) or
> maybe i would not output such a strong statement.
>
> Regards,
> --
> Sandro Tosi (aka morph, morpheus, matrixhasu)
> My website: http://matrixhasu.altervista.org/
> Me at Debian: http://wiki.debian.org/SandroTosi
>
>
> ------------------------------------------------------------------------------
> Dive into the World of Parallel Programming The Go Parallel Website,
> sponsored
> by Intel and developed in partnership with Slashdot Media, is your hub for
> all
> things parallel software development, from weekly thought leadership blogs
> to
> news, videos, case studies, tutorials and more. Take a look and join the
> conversation now. http://goparallel.sourceforge.net/
> _______________________________________________
> Matplotlib-devel mailing list
> Mat...@li...
> https://lists.sourceforge.net/lists/listinfo/matplotlib-devel
>

On Wed, Mar 18, 2015 at 3:33 PM, Benjamin Root <ben...@ou...> wrote:
> I would just use pip. Ubuntu/Debian has really messed up the python
> environment in more ways than one.
<debian developer hat=on/> ehm... either you back up your claims
(and/or try to understand the reasons behind the distro needs) or
maybe i would not output such a strong statement.
Regards,
-- 
Sandro Tosi (aka morph, morpheus, matrixhasu)
My website: http://matrixhasu.altervista.org/
Me at Debian: http://wiki.debian.org/SandroTosi

pip still thinks I have the latest. I think it's a question of how to force the matplotlib setup.py to use actually it.
Keith
 
kbriggs:~/Downloads/matplotlib-1.4.3> sudo pip3 install setuptools --upgrade 
Requirement already up-to-date: setuptools in /usr/local/lib/python3.4/dist-packages/setuptools-14.3-py3.4.egg
Cleaning up...
________________________________________
From: ben...@gm... [ben...@gm...] On Behalf Of Benjamin Root [ben...@ou...]
Sent: 18 March 2015 15:33
To: Briggs,KM,Keith,TUB2 R
Cc: matplotlib development list
Subject: Re: [matplotlib-devel] 1.4.3 does not build on Ubuntu 14 with python3
I would just use pip. Ubuntu/Debian has really messed up the python environment in more ways than one.
On Wed, Mar 18, 2015 at 11:08 AM, keithbriggs <kei...@bt...<mailto:kei...@bt...>> wrote:
The Ubuntu package manager tells me it is up to date.
If I download setuptools-14.3 and install, it goes into /usr/local/lib/ and
doesn't get used.
How do I force it to be used?
Or am I supposed to override the Ubuntu package manager?
Keith

I would just use pip. Ubuntu/Debian has really messed up the python
environment in more ways than one.
On Wed, Mar 18, 2015 at 11:08 AM, keithbriggs <kei...@bt...> wrote:
> The Ubuntu package manager tells me it is up to date.
> If I download setuptools-14.3 and install, it goes into /usr/local/lib/ and
> doesn't get used.
> How do I force it to be used?
> Or am I supposed to override the Ubuntu package manager?
> Keith
>
>
>
> --
> View this message in context:
> http://matplotlib.1069221.n5.nabble.com/1-4-3-does-not-build-on-Ubuntu-14-with-python3-tp45198p45201.html
> Sent from the matplotlib - devel mailing list archive at Nabble.com.
>
>
> ------------------------------------------------------------------------------
> Dive into the World of Parallel Programming The Go Parallel Website,
> sponsored
> by Intel and developed in partnership with Slashdot Media, is your hub for
> all
> things parallel software development, from weekly thought leadership blogs
> to
> news, videos, case studies, tutorials and more. Take a look and join the
> conversation now. http://goparallel.sourceforge.net/
> _______________________________________________
> Matplotlib-devel mailing list
> Mat...@li...
> https://lists.sourceforge.net/lists/listinfo/matplotlib-devel
>

The Ubuntu package manager tells me it is up to date.
If I download setuptools-14.3 and install, it goes into /usr/local/lib/ and
doesn't get used.
How do I force it to be used?
Or am I supposed to override the Ubuntu package manager?
Keith
--
View this message in context: http://matplotlib.1069221.n5.nabble.com/1-4-3-does-not-build-on-Ubuntu-14-with-python3-tp45198p45201.html
Sent from the matplotlib - devel mailing list archive at Nabble.com.

Also see https://github.com/matplotlib/matplotlib/issues/3889
On Wed, Mar 18, 2015 at 10:22 AM Benjamin Root <ben...@ou...> wrote:
> Please update your install of setuptools.
>
> On Wed, Mar 18, 2015 at 10:01 AM, <kei...@bt...> wrote:
>
>> kbriggs:~/Downloads/matplotlib-1.4.3> python3 setup.py build
>> ============================================================================
>> Edit setup.cfg to change the build options
>>
>> BUILDING MATPLOTLIB
>> matplotlib: yes [1.4.3]
>> python: yes [3.4.0 (default, Apr 11 2014, 13:05:11) [GCC
>> 4.8.2]]
>> platform: yes [linux]
>>
>> REQUIRED DEPENDENCIES AND EXTENSIONS
>> numpy: yes [version 1.8.2]
>> six: yes [using six version 1.5.2]
>> dateutil: yes [dateutil was not found. It is required for
>> date
>> axis support. pip/easy_install may attempt to
>> install it after matplotlib.]
>> pytz: yes [pytz was not found. pip will attempt to
>> install
>> it after matplotlib.]
>> tornado: yes [tornado was not found. It is required for the
>> WebAgg backend. pip/easy_install may attempt to
>> install it after matplotlib.]
>> pyparsing: yes [using pyparsing version 2.0.1]
>> pycxx: yes [Official versions of PyCXX are not compatible
>> with matplotlib on Python 3.x, since they lack
>> support for the buffer object. Using local copy]
>> libagg: yes [pkg-config information for 'libagg' could not
>> be found. Using local copy.] Traceback (most
>> recent call last):
>> File "setup.py", line 155, in <module>
>> result = package.check()
>> File "/home/kbriggs/Downloads/matplotlib-1.4.3/setupext.py", line 961,
>> in check
>> min_version='2.3', version=version)
>> File "/home/kbriggs/Downloads/matplotlib-1.4.3/setupext.py", line 445,
>> in _check_for_pkg_config
>> if (not is_min_version(version, min_version)):
>> File "/home/kbriggs/Downloads/matplotlib-1.4.3/setupext.py", line 173,
>> in is_min_version
>> return found_version >= expected_version
>> File "/usr/lib/python3.4/distutils/version.py", line 76, in __ge__
>> c = self._cmp(other)
>> File "/usr/lib/python3.4/distutils/version.py", line 343, in _cmp
>> if self.version < other.version:
>> TypeError: unorderable types: str() < int()
>>
>>
>>
>> ------------------------------------------------------------------------------
>> Dive into the World of Parallel Programming The Go Parallel Website,
>> sponsored
>> by Intel and developed in partnership with Slashdot Media, is your hub
>> for all
>> things parallel software development, from weekly thought leadership
>> blogs to
>> news, videos, case studies, tutorials and more. Take a look and join the
>> conversation now. http://goparallel.sourceforge.net/
>> _______________________________________________
>> Matplotlib-devel mailing list
>> Mat...@li...
>> https://lists.sourceforge.net/lists/listinfo/matplotlib-devel
>>
>
> ------------------------------------------------------------
> ------------------
> Dive into the World of Parallel Programming The Go Parallel Website,
> sponsored
> by Intel and developed in partnership with Slashdot Media, is your hub for
> all
> things parallel software development, from weekly thought leadership blogs
> to
> news, videos, case studies, tutorials and more. Take a look and join the
> conversation now. http://goparallel.sourceforge.net/
> _______________________________________________
> Matplotlib-devel mailing list
> Mat...@li...
> https://lists.sourceforge.net/lists/listinfo/matplotlib-devel
>

Please update your install of setuptools.
On Wed, Mar 18, 2015 at 10:01 AM, <kei...@bt...> wrote:
> kbriggs:~/Downloads/matplotlib-1.4.3> python3 setup.py build
> ============================================================================
> Edit setup.cfg to change the build options
>
> BUILDING MATPLOTLIB
> matplotlib: yes [1.4.3]
> python: yes [3.4.0 (default, Apr 11 2014, 13:05:11) [GCC
> 4.8.2]]
> platform: yes [linux]
>
> REQUIRED DEPENDENCIES AND EXTENSIONS
> numpy: yes [version 1.8.2]
> six: yes [using six version 1.5.2]
> dateutil: yes [dateutil was not found. It is required for
> date
> axis support. pip/easy_install may attempt to
> install it after matplotlib.]
> pytz: yes [pytz was not found. pip will attempt to
> install
> it after matplotlib.]
> tornado: yes [tornado was not found. It is required for the
> WebAgg backend. pip/easy_install may attempt to
> install it after matplotlib.]
> pyparsing: yes [using pyparsing version 2.0.1]
> pycxx: yes [Official versions of PyCXX are not compatible
> with matplotlib on Python 3.x, since they lack
> support for the buffer object. Using local copy]
> libagg: yes [pkg-config information for 'libagg' could not
> be found. Using local copy.] Traceback (most
> recent call last):
> File "setup.py", line 155, in <module>
> result = package.check()
> File "/home/kbriggs/Downloads/matplotlib-1.4.3/setupext.py", line 961,
> in check
> min_version='2.3', version=version)
> File "/home/kbriggs/Downloads/matplotlib-1.4.3/setupext.py", line 445,
> in _check_for_pkg_config
> if (not is_min_version(version, min_version)):
> File "/home/kbriggs/Downloads/matplotlib-1.4.3/setupext.py", line 173,
> in is_min_version
> return found_version >= expected_version
> File "/usr/lib/python3.4/distutils/version.py", line 76, in __ge__
> c = self._cmp(other)
> File "/usr/lib/python3.4/distutils/version.py", line 343, in _cmp
> if self.version < other.version:
> TypeError: unorderable types: str() < int()
>
>
>
> ------------------------------------------------------------------------------
> Dive into the World of Parallel Programming The Go Parallel Website,
> sponsored
> by Intel and developed in partnership with Slashdot Media, is your hub for
> all
> things parallel software development, from weekly thought leadership blogs
> to
> news, videos, case studies, tutorials and more. Take a look and join the
> conversation now. http://goparallel.sourceforge.net/
> _______________________________________________
> Matplotlib-devel mailing list
> Mat...@li...
> https://lists.sourceforge.net/lists/listinfo/matplotlib-devel
>

kbriggs:~/Downloads/matplotlib-1.4.3> python3 setup.py build ============================================================================
Edit setup.cfg to change the build options
BUILDING MATPLOTLIB
 matplotlib: yes [1.4.3]
 python: yes [3.4.0 (default, Apr 11 2014, 13:05:11) [GCC
 4.8.2]]
 platform: yes [linux]
REQUIRED DEPENDENCIES AND EXTENSIONS
 numpy: yes [version 1.8.2]
 six: yes [using six version 1.5.2]
 dateutil: yes [dateutil was not found. It is required for date
 axis support. pip/easy_install may attempt to
 install it after matplotlib.]
 pytz: yes [pytz was not found. pip will attempt to install
 it after matplotlib.]
 tornado: yes [tornado was not found. It is required for the
 WebAgg backend. pip/easy_install may attempt to
 install it after matplotlib.]
 pyparsing: yes [using pyparsing version 2.0.1]
 pycxx: yes [Official versions of PyCXX are not compatible
 with matplotlib on Python 3.x, since they lack
 support for the buffer object. Using local copy]
 libagg: yes [pkg-config information for 'libagg' could not
 be found. Using local copy.] Traceback (most recent call last):
 File "setup.py", line 155, in <module>
 result = package.check()
 File "/home/kbriggs/Downloads/matplotlib-1.4.3/setupext.py", line 961, in check
 min_version='2.3', version=version)
 File "/home/kbriggs/Downloads/matplotlib-1.4.3/setupext.py", line 445, in _check_for_pkg_config
 if (not is_min_version(version, min_version)):
 File "/home/kbriggs/Downloads/matplotlib-1.4.3/setupext.py", line 173, in is_min_version
 return found_version >= expected_version
 File "/usr/lib/python3.4/distutils/version.py", line 76, in __ge__
 c = self._cmp(other)
 File "/usr/lib/python3.4/distutils/version.py", line 343, in _cmp
 if self.version < other.version:
TypeError: unorderable types: str() < int()

I have opened a PR to document this discussion. It is meant to provide a
permanent record of the thought process leading up to color map and to
serve as a tool in making the finial decision.
https://github.com/matplotlib/matplotlib/pull/4238
On Mon, Mar 2, 2015 at 6:32 AM jni <jni...@gm...> wrote:
> Hi Pierre,
>
> Could you please elaborate a bit on this
>> usecase. I was thinking, naively, that when plotting a grayscale image,
>> one would simply used a gray colormap.
>>
>
> Using a colormap with hue and saturation gives you better contrast than
> pure grayscale. For natural images, that is, photographs of human-scale
> objects, indeed grayscale is a good choice, because that is how we are used
> to looking at those images. But for looking at physical quantities, for
> example, using a colormap with hue and saturation as well as lightness is
> useful. Here are some examples:
> http://www.gnuplotting.org/color-maps-from-colorbrewer/
> https://www.mrao.cam.ac.uk/~dag/CUBEHELIX/
>
> See also a "boundary probability map" for a natural image here (panel B,
> top right):
>
> http://www.frontiersin.org/files/Articles/74212/fninf-08-00034-r2/image_m/fninf-08-00034-g001.jpg
> Having the colormap makes it easier to place the intermediate levels of
> the probability map.
>
> Again, restricting the lightness range for these maps would be
> problematic, to say the least.
>
> Juan.
>
> ------------------------------
> View this message in context: Re: release strategy and the color
> revolution
> <http://matplotlib.1069221.n5.nabble.com/release-strategy-and-the-color-revolution-tp44929p45030.html>
> Sent from the matplotlib - devel mailing list archive
> <http://matplotlib.1069221.n5.nabble.com/matplotlib-devel-f28077.html> at
> Nabble.com.
> ------------------------------------------------------------
> ------------------
> Dive into the World of Parallel Programming The Go Parallel Website,
> sponsored
> by Intel and developed in partnership with Slashdot Media, is your hub for
> all
> things parallel software development, from weekly thought leadership blogs
> to
> news, videos, case studies, tutorials and more. Take a look and join the
> conversation now. http://goparallel.sourceforge.net/
> _______________________________________________
> Matplotlib-devel mailing list
> Mat...@li...
> https://lists.sourceforge.net/lists/listinfo/matplotlib-devel
>