Alpha keyword not working with contourf

Eric_Firing1 · February 11, 2012, 8:48pm

Thanks for replying Eric.

(You're welcome. But please keep replies on the list when they are potentially useful to others, as yours is.)

Here is my minimal script -

import matplotlib

frompylab import*

import numpy as np

importos

metals=np.arange(-3.0, 1.1, 0.1)

U=np.arange(-6.0, 0.25, 0.25)

o3=np.zeros([25,41])

plt.contourf(metals, U, o3, levels=[o3col-nsig*o3sig], alpha=0.20,
colors='blue')

plt.savefig("CoutourPlot.ps")

The Postscript language does not support transparency, so transparency is lost when you use the mpl postscript backend. The way to work around this, if you really need to end up with a postscript file, is to save the file as pdf, and then use a converter program to render that as postscript. (It might be a "print-to-file" or "save-as" option on your pdf display program, for example. I don't know what is typically available on a Mac, but Macs have long been pdf-friendly. In general, the need for ps files has been diminishing. I almost never generate them any more.)

Eric

···

On 02/11/2012 10:30 AM, Alexa Villaume wrote:

It plots fine except except there is no transparency in colors of the
contour, it's like alpha is always set to 1.0

On Sat, Feb 11, 2012 at 8:23 AM, Eric Firing <efiring@...202... > <mailto:efiring@…202…>> wrote:

    On 02/11/2012 07:40 AM, Alexa Villaume wrote:
     > Hi All,
     >
     > I'm just getting started with matplotlib. I'm trying to make a
    contour
     > plot using contourf and have the different paths colored and semi
     > transparent but the alpha keyword doesn't seem to do anything. I
    googled
     > around and found that other people have had this same problem but I
     > didn't find a solution. Here's what I'm doing -
     >
     > contourf(metals, U, o3, levels=[o3col-nsig*o3sig, o3col+nsig*o3sig],
     > alpha=0.20, colors='magenta')
     >
     > contourf(metals, U, o2, levels=[o2col-nsig*o2sig, o2col+nsig*o2sig],
     > alpha=0.20, colors='blue')
     >
     > contourf(metals, U, c3, levels=[c3col-nsig*c3sig, c3col+nsig*c3sig],
     > alpha=0.20, colors='green')
     >
     > To plot them all together.
     >
     > For reference my machine is running Mac OS X 10.6.8 and my version of
     > matplotlib is 1.2.x
     >
     > Thanks!

    Please provide a minimal but complete script that illustrates the
    problem; describe what it is in the output that does not match your
    expectations. e.g.,

    import numpy as np
    import matplotlib.pyplot as plt
    z = np.arange(20).reshape(4,5)
    plt.contourf(z, levels=[2,3], alpha=0.2, colors='magenta')
    plt.savefig("testcontourf_alpha.png")

    which produces a pale magenta stripe, as expected, with mpl from github
    master.

    Eric

Alexa · February 12, 2012, 8:53am

(Sorry about not replying to list before. I usually have to be pretty vigilant about not clicking “Reply-All”)

I made my figure a pdf and the transparency works fine. So thanks! Just another quick question is it the Postscript language itself that doesn’t support transparency or the way mpl handles postscript files?

···

On Sat, Feb 11, 2012 at 10:48 AM, Eric Firing <efiring@…3178…> wrote:

On 02/11/2012 10:30 AM, Alexa Villaume wrote:

Thanks for replying Eric.

(You’re welcome. But please keep replies on the list when they are potentially useful to others, as yours is.)

Here is my minimal script -

import matplotlib

frompylab import*

import numpy as np

importos

metals=np.arange(-3.0, 1.1, 0.1)

U=np.arange(-6.0, 0.25, 0.25)

o3=np.zeros([25,41])

plt.contourf(metals, U, o3, levels=[o3col-nsig*o3sig], alpha=0.20,

colors=‘blue’)

plt.savefig(“CoutourPlot.ps”)

The Postscript language does not support transparency, so transparency is lost when you use the mpl postscript backend. The way to work around this, if you really need to end up with a postscript file, is to save the file as pdf, and then use a converter program to render that as postscript. (It might be a “print-to-file” or “save-as” option on your pdf display program, for example. I don’t know what is typically available on a Mac, but Macs have long been pdf-friendly. In general, the need for ps files has been diminishing. I almost never generate them any more.)

Eric
It plots fine except except there is no transparency in colors of the

contour, it’s like alpha is always set to 1.0

On Sat, Feb 11, 2012 at 8:23 AM, Eric Firing <efiring@…202… > > mailto:efiring@...120.....202...> wrote:
On 02/11/2012 07:40 AM, Alexa Villaume wrote:

 > Hi All,

 >

 > I'm just getting started with matplotlib. I'm trying to make a

contour

 > plot using contourf and have the different paths colored and semi

 > transparent but the alpha keyword doesn't seem to do anything. I

googled

 > around and found that other people have had this same problem but I

 > didn't find a solution. Here's what I'm doing -

 >

 > contourf(metals, U, o3, levels=[o3col-nsig*o3sig, o3col+nsig*o3sig],

 > alpha=0.20, colors='magenta')

 >

 > contourf(metals, U, o2, levels=[o2col-nsig*o2sig, o2col+nsig*o2sig],

 > alpha=0.20, colors='blue')

 >

 > contourf(metals, U, c3, levels=[c3col-nsig*c3sig, c3col+nsig*c3sig],

 > alpha=0.20, colors='green')

 >

 >

 > To plot them all together.

 >

 > For reference my machine is running Mac OS X 10.6.8 and my version of

 > matplotlib is 1.2.x

 >

 > Thanks!



Please provide a minimal but complete script that illustrates the

problem; describe what it is in the output that does not match your

expectations.  e.g.,



import numpy as np

import matplotlib.pyplot as plt

z = np.arange(20).reshape(4,5)

plt.contourf(z, levels=[2,3], alpha=0.2, colors='magenta')

plt.savefig("testcontourf_alpha.png")



which produces a pale magenta stripe, as expected, with mpl from github

master.



Eric

Eric_Firing1 · February 12, 2012, 5:46pm

(Sorry about not replying to list before. I usually have to be pretty
vigilant about not clicking "Reply-All")

I made my figure a pdf and the transparency works fine. So thanks! Just
another quick question is it the Postscript language itself that doesn't
support transparency or the way mpl handles postscript files?

It is the language itself.

Eric

···

On 02/11/2012 10:53 PM, Alexa Villaume wrote:

On Sat, Feb 11, 2012 at 10:48 AM, Eric Firing <efiring@...202... > <mailto:efiring@…202…>> wrote:

    On 02/11/2012 10:30 AM, Alexa Villaume wrote:

        Thanks for replying Eric.

    (You're welcome. But please keep replies on the list when they are
    potentially useful to others, as yours is.)

        Here is my minimal script -

        import matplotlib

        frompylab import*

        import numpy as np

        importos

        metals=np.arange(-3.0, 1.1, 0.1)

        U=np.arange(-6.0, 0.25, 0.25)

        o3=np.zeros([25,41])

        plt.contourf(metals, U, o3, levels=[o3col-nsig*o3sig], alpha=0.20,
        colors='blue')

        plt.savefig("CoutourPlot.ps")

    The Postscript language does not support transparency, so
    transparency is lost when you use the mpl postscript backend. The
    way to work around this, if you really need to end up with a
    postscript file, is to save the file as pdf, and then use a
    converter program to render that as postscript. (It might be a
    "print-to-file" or "save-as" option on your pdf display program, for
    example. I don't know what is typically available on a Mac, but
    Macs have long been pdf-friendly. In general, the need for ps files
    has been diminishing. I almost never generate them any more.)

    Eric

        It plots fine except except there is no transparency in colors
        of the
        contour, it's like alpha is always set to 1.0

        On Sat, Feb 11, 2012 at 8:23 AM, Eric Firing <efiring@...202... > <mailto:efiring@…202…> > <mailto:efiring@…202…>> wrote:

            On 02/11/2012 07:40 AM, Alexa Villaume wrote:
         > Hi All,
         >
         > I'm just getting started with matplotlib. I'm trying to make a
            contour
         > plot using contourf and have the different paths colored and semi
         > transparent but the alpha keyword doesn't seem to do anything. I
            googled
         > around and found that other people have had this same problem
        but I
         > didn't find a solution. Here's what I'm doing -
         >
         > contourf(metals, U, o3, levels=[o3col-nsig*o3sig,
        o3col+nsig*o3sig],
         > alpha=0.20, colors='magenta')
         >
         > contourf(metals, U, o2, levels=[o2col-nsig*o2sig,
        o2col+nsig*o2sig],
         > alpha=0.20, colors='blue')
         >
         > contourf(metals, U, c3, levels=[c3col-nsig*c3sig,
        c3col+nsig*c3sig],
         > alpha=0.20, colors='green')
         >
         > To plot them all together.
         >
         > For reference my machine is running Mac OS X 10.6.8 and my
        version of
         > matplotlib is 1.2.x
         >
         > Thanks!

            Please provide a minimal but complete script that
        illustrates the
            problem; describe what it is in the output that does not
        match your
            expectations. e.g.,

            import numpy as np
            import matplotlib.pyplot as plt
            z = np.arange(20).reshape(4,5)
            plt.contourf(z, levels=[2,3], alpha=0.2, colors='magenta')
            plt.savefig("testcontourf___alpha.png")

            which produces a pale magenta stripe, as expected, with mpl
        from github
            master.

            Eric