matplotlib-users Mailing List for matplotlib

Skip,

Here are some lines from an application I have written.

from matplotlib.backends.backend_gtkagg import FigureCanvasGTKAgg

In the setup:
    self.canvas = FigureCanvasGTKAgg(self.figure)
    self.canvas.set_size_request(600,600)
    self.canvas.show()
    #Pack the canvas in a parent container
    self.vbox0.pack_start(self.canvas, True, True)

At the event where I want the plot to update:
    self.canvas.draw()

For a faster responding application, be sure to check out the matplotlib animation examples with blitting, as I built my application without that knowledge, and haven't found the time to go back and fix it...

-Sterling

On Oct 11, 2013, at 9:25AM, Skip Montanaro wrote:

> I want to use matplotlib as a component of a larger, event-driven GTK
> app. That means pylab.show is (I think) not the way to go, as it
> starts up its own event loop which doesn't return. I've tried to clear
> and plot in my event handler, but my plot is never displayed.
> 
> My initialization code looks like this:
> 
>        matplotlib.use("GtkAgg")
>        ....
>        self.figure = matplotlib.figure.Figure()
>        self.plot = self.figure.add_subplot(111)
>        self.plot.set_axisbelow(True)
>        self.figure.tight_layout()
> 
> My event handler computes a new set of points (about a dozen x/y
> pairs) and plots them:
> 
>        points = ... generate a list of (x, y) tuples ...
>        print points
>        self.plot.clear()
>        self.plot.plot([x for x, y in points], [y for x, y in points])
> 
> Every time my event handler is called, it prints the points (I see
> them in my xterm), but the plot is never drawn.
> 
> I looked at this example:
> 
> http://matplotlib.org/examples/user_interfaces/embedding_in_gtk2.html
> 
> but it does something with a key press handler that seems very
> artificial, and not at all like how my application will interact with
> its environment.  Can someone point me to the correct spot in the
> documentation or some examples that don't require the user to type at
> the application?
> 
> Thanks,
> 
> Skip
> 
> ------------------------------------------------------------------------------
> October Webinars: Code for Performance
> Free Intel webinars can help you accelerate application performance.
> Explore tips for MPI, OpenMP, advanced profiling, and more. Get the most from 
> the latest Intel processors and coprocessors. See abstracts and register >
> http://pubads.g.doubleclick.net/gampad/clk?id=60134071&iu=/4140/ostg.clktrk
> _______________________________________________
> Matplotlib-users mailing list
> Mat...@li...
> https://lists.sourceforge.net/lists/listinfo/matplotlib-users

I want to use matplotlib as a component of a larger, event-driven GTK
app. That means pylab.show is (I think) not the way to go, as it
starts up its own event loop which doesn't return. I've tried to clear
and plot in my event handler, but my plot is never displayed.

My initialization code looks like this:

        matplotlib.use("GtkAgg")
        ....
        self.figure = matplotlib.figure.Figure()
        self.plot = self.figure.add_subplot(111)
        self.plot.set_axisbelow(True)
        self.figure.tight_layout()

My event handler computes a new set of points (about a dozen x/y
pairs) and plots them:

        points = ... generate a list of (x, y) tuples ...
        print points
        self.plot.clear()
        self.plot.plot([x for x, y in points], [y for x, y in points])

Every time my event handler is called, it prints the points (I see
them in my xterm), but the plot is never drawn.

I looked at this example:

http://matplotlib.org/examples/user_interfaces/embedding_in_gtk2.html

but it does something with a key press handler that seems very
artificial, and not at all like how my application will interact with
its environment.  Can someone point me to the correct spot in the
documentation or some examples that don't require the user to type at
the application?

Thanks,

Skip

Nils,

I tried to run your example, but there are some variables which are undefined.  Can you post a self contained revision of your example?

-Sterling

On Oct 11, 2013, at 1:34AM, Nils Wagner wrote:

> plt.colorbar(scalarMap,ax=ax) results in
> 
> cm.py", line 309, in autoscale_None
>     raise TypeError('You must first set_array for mappable')
> TypeError: You must first set_array for mappable
> 
> Nils
> 
> 
> 
> On Fri, Oct 11, 2013 at 9:51 AM, Eric Firing <ef...@ha...> wrote:
> On 2013/10/10 8:52 PM, Nils Wagner wrote:
> > Hi all,
> >
> > I tried to add a colorbar to a bar plot
> >
> > coolwarm = cm =  plt.get_cmap('coolwarm')
> > values = range(100)
> > cNorm  = colors.Normalize(vmin=0, vmax=values[-1])
> > scalarMap = cmx.ScalarMappable(norm=cNorm, cmap=coolwarm)
> > colours = []
> > for value in values:
> >      colorVal = scalarMap.to_rgba(value)
> >      colours.append(colorVal)
> >
> > fig = plt.figure()
> > ax  = fig.add_subplot(111,projection='3d')
> > hist,bin_edges = np.histogram(efratio,bins=100,range=(0.,1.),density=False)
> > width = 0.7*(bin_edges[1]-bin_edges[0])
> > center = (bin_edges[:-1]+bin_edges[1:])/2
> > heatmap = ax.bar(center, hist, zs=z, zdir='y', align = 'center', width =
> > width,color=colours)
> > plt.colorbar(heatmap)
> >
> >
> >
> >
> >
> >      mappable.autoscale_None() # Ensure mappable.norm.vmin, vmax
> > AttributeError: 'BarContainer' object has no attribute 'autoscale_None'
> 
> This is because it is not an instance of ScalarMappable, which is what
> colorbar() requires as its argument.
> >
> > How can I fix the problem ?
> 
> Use scalarMap as the argument instead of heatmap.  I think you will need
> to provide either the cax or the ax kwarg in addition.
> 
> examples/api/colorbar_only.py might also be helpful.
> 
> Eric
> >
> > Nils
> >
> >
> >
> > ------------------------------------------------------------------------------
> > October Webinars: Code for Performance
> > Free Intel webinars can help you accelerate application performance.
> > Explore tips for MPI, OpenMP, advanced profiling, and more. Get the most from
> > the latest Intel processors and coprocessors. See abstracts and register >
> > http://pubads.g.doubleclick.net/gampad/clk?id=60134071&iu=/4140/ostg.clktrk
> >
> >
> >
> > _______________________________________________
> > Matplotlib-users mailing list
> > Mat...@li...
> > https://lists.sourceforge.net/lists/listinfo/matplotlib-users
> >
> 
> 
> ------------------------------------------------------------------------------
> October Webinars: Code for Performance
> Free Intel webinars can help you accelerate application performance.
> Explore tips for MPI, OpenMP, advanced profiling, and more. Get the most from
> the latest Intel processors and coprocessors. See abstracts and register >
> http://pubads.g.doubleclick.net/gampad/clk?id=60134071&iu=/4140/ostg.clktrk
> _______________________________________________
> Matplotlib-users mailing list
> Mat...@li...
> https://lists.sourceforge.net/lists/listinfo/matplotlib-users
> 
> ------------------------------------------------------------------------------
> October Webinars: Code for Performance
> Free Intel webinars can help you accelerate application performance.
> Explore tips for MPI, OpenMP, advanced profiling, and more. Get the most from 
> the latest Intel processors and coprocessors. See abstracts and register >
> http://pubads.g.doubleclick.net/gampad/clk?id=60134071&iu=/4140/ostg.clktrk_______________________________________________
> Matplotlib-users mailing list
> Mat...@li...
> https://lists.sourceforge.net/lists/listinfo/matplotlib-users

plt.colorbar(scalarMap,ax=ax) results in

cm.py", line 309, in autoscale_None
    raise TypeError('You must first set_array for mappable')
TypeError: You must first set_array for mappable

Nils



On Fri, Oct 11, 2013 at 9:51 AM, Eric Firing <ef...@ha...> wrote:

> On 2013/10/10 8:52 PM, Nils Wagner wrote:
> > Hi all,
> >
> > I tried to add a colorbar to a bar plot
> >
> > coolwarm = cm =  plt.get_cmap('coolwarm')
> > values = range(100)
> > cNorm  = colors.Normalize(vmin=0, vmax=values[-1])
> > scalarMap = cmx.ScalarMappable(norm=cNorm, cmap=coolwarm)
> > colours = []
> > for value in values:
> >      colorVal = scalarMap.to_rgba(value)
> >      colours.append(colorVal)
> >
> > fig = plt.figure()
> > ax  = fig.add_subplot(111,projection='3d')
> > hist,bin_edges =
> np.histogram(efratio,bins=100,range=(0.,1.),density=False)
> > width = 0.7*(bin_edges[1]-bin_edges[0])
> > center = (bin_edges[:-1]+bin_edges[1:])/2
> > heatmap = ax.bar(center, hist, zs=z, zdir='y', align = 'center', width =
> > width,color=colours)
> > plt.colorbar(heatmap)
> >
> >
> >
> >
> >
> >      mappable.autoscale_None() # Ensure mappable.norm.vmin, vmax
> > AttributeError: 'BarContainer' object has no attribute 'autoscale_None'
>
> This is because it is not an instance of ScalarMappable, which is what
> colorbar() requires as its argument.
> >
> > How can I fix the problem ?
>
> Use scalarMap as the argument instead of heatmap.  I think you will need
> to provide either the cax or the ax kwarg in addition.
>
> examples/api/colorbar_only.py might also be helpful.
>
> Eric
> >
> > Nils
> >
> >
> >
> >
> ------------------------------------------------------------------------------
> > October Webinars: Code for Performance
> > Free Intel webinars can help you accelerate application performance.
> > Explore tips for MPI, OpenMP, advanced profiling, and more. Get the most
> from
> > the latest Intel processors and coprocessors. See abstracts and register
> >
> >
> http://pubads.g.doubleclick.net/gampad/clk?id=60134071&iu=/4140/ostg.clktrk
> >
> >
> >
> > _______________________________________________
> > Matplotlib-users mailing list
> > Mat...@li...
> > https://lists.sourceforge.net/lists/listinfo/matplotlib-users
> >
>
>
>
> ------------------------------------------------------------------------------
> October Webinars: Code for Performance
> Free Intel webinars can help you accelerate application performance.
> Explore tips for MPI, OpenMP, advanced profiling, and more. Get the most
> from
> the latest Intel processors and coprocessors. See abstracts and register >
> http://pubads.g.doubleclick.net/gampad/clk?id=60134071&iu=/4140/ostg.clktrk
> _______________________________________________
> Matplotlib-users mailing list
> Mat...@li...
> https://lists.sourceforge.net/lists/listinfo/matplotlib-users
>

On 2013/10/10 8:52 PM, Nils Wagner wrote:
> Hi all,
>
> I tried to add a colorbar to a bar plot
>
> coolwarm = cm =  plt.get_cmap('coolwarm')
> values = range(100)
> cNorm  = colors.Normalize(vmin=0, vmax=values[-1])
> scalarMap = cmx.ScalarMappable(norm=cNorm, cmap=coolwarm)
> colours = []
> for value in values:
>      colorVal = scalarMap.to_rgba(value)
>      colours.append(colorVal)
>
> fig = plt.figure()
> ax  = fig.add_subplot(111,projection='3d')
> hist,bin_edges = np.histogram(efratio,bins=100,range=(0.,1.),density=False)
> width = 0.7*(bin_edges[1]-bin_edges[0])
> center = (bin_edges[:-1]+bin_edges[1:])/2
> heatmap = ax.bar(center, hist, zs=z, zdir='y', align = 'center', width =
> width,color=colours)
> plt.colorbar(heatmap)
>
>
>
>
>
>      mappable.autoscale_None() # Ensure mappable.norm.vmin, vmax
> AttributeError: 'BarContainer' object has no attribute 'autoscale_None'

This is because it is not an instance of ScalarMappable, which is what 
colorbar() requires as its argument.
>
> How can I fix the problem ?

Use scalarMap as the argument instead of heatmap.  I think you will need 
to provide either the cax or the ax kwarg in addition.

examples/api/colorbar_only.py might also be helpful.

Eric
>
> Nils
>
>
>
> ------------------------------------------------------------------------------
> October Webinars: Code for Performance
> Free Intel webinars can help you accelerate application performance.
> Explore tips for MPI, OpenMP, advanced profiling, and more. Get the most from
> the latest Intel processors and coprocessors. See abstracts and register >
> http://pubads.g.doubleclick.net/gampad/clk?id=60134071&iu=/4140/ostg.clktrk
>
>
>
> _______________________________________________
> Matplotlib-users mailing list
> Mat...@li...
> https://lists.sourceforge.net/lists/listinfo/matplotlib-users
>

Hi all,

I tried to add a colorbar to a bar plot

coolwarm = cm =  plt.get_cmap('coolwarm')
values = range(100)
cNorm  = colors.Normalize(vmin=0, vmax=values[-1])
scalarMap = cmx.ScalarMappable(norm=cNorm, cmap=coolwarm)
colours = []
for value in values:
    colorVal = scalarMap.to_rgba(value)
    colours.append(colorVal)

fig = plt.figure()
ax  = fig.add_subplot(111,projection='3d')
hist,bin_edges = np.histogram(efratio,bins=100,range=(0.,1.),density=False)
width = 0.7*(bin_edges[1]-bin_edges[0])
center = (bin_edges[:-1]+bin_edges[1:])/2
heatmap = ax.bar(center, hist, zs=z, zdir='y', align = 'center', width =
width,color=colours)
plt.colorbar(heatmap)





    mappable.autoscale_None() # Ensure mappable.norm.vmin, vmax
AttributeError: 'BarContainer' object has no attribute 'autoscale_None'

How can I fix the problem ?

Nils