jpg not supported by matplotlib backend

[jankatins]

Just got an error after enabling the jpg matplotlib format:

%matplotlib inline
from IPython.display import set_matplotlib_formats
set_matplotlib_formats('png', 'jpg')

ValueError                                Traceback (most recent call last)
C:\portabel\miniconda\envs\ipython\lib\site-packages\IPython\core\formatters.pyc in __call__(self, obj)
    333                 pass
    334             else:
--> 335                 return printer(obj)
    336             # Finally look for special method names
    337             method = _safe_get_formatter_method(obj, self.print_method)

C:\portabel\miniconda\envs\ipython\lib\site-packages\IPython\core\pylabtools.pyc in <lambda>(fig)
    209         png_formatter.for_type(Figure, lambda fig: retina_figure(fig, **kwargs))
    210     if 'jpg' in formats or 'jpeg' in formats:
--> 211         jpg_formatter.for_type(Figure, lambda fig: print_figure(fig, 'jpg', **kwargs))
    212     if 'svg' in formats:
    213         svg_formatter.for_type(Figure, lambda fig: print_figure(fig, 'svg', **kwargs))

C:\portabel\miniconda\envs\ipython\lib\site-packages\IPython\core\pylabtools.pyc in print_figure(fig, fmt, bbox_inches, **kwargs)
    115     
    116     bytes_io = BytesIO()
--> 117     fig.canvas.print_figure(bytes_io, **kw)
    118     data = bytes_io.getvalue()
    119     if fmt == 'svg':

C:\portabel\miniconda\envs\ipython\lib\site-packages\matplotlib\backend_bases.pyc in print_figure(self, filename, dpi, facecolor, edgecolor, orientation, format, **kwargs)
   2117 
   2118         # get canvas object and print method for format
-> 2119         canvas = self._get_output_canvas(format)
   2120         print_method = getattr(canvas, 'print_%s' % format)
   2121 

C:\portabel\miniconda\envs\ipython\lib\site-packages\matplotlib\backend_bases.pyc in _get_output_canvas(self, format)
   2059         raise ValueError('Format "%s" is not supported.\n'
   2060                          'Supported formats: '
-> 2061                          '%s.' % (format, ', '.join(formats)))
   2062 
   2063     def print_figure(self, filename, dpi=None, facecolor='w', edgecolor='w',

ValueError: Format "jpg" is not supported.
Supported formats: eps, pdf, pgf, png, ps, raw, rgba, svg, svgz.

IPython.core.pylabtools has this here:

    supported = {'png', 'png2x', 'retina', 'jpg', 'jpeg', 'svg', 'pdf'}

So I suspect that something is wrong here...

This is IPython 3.0, matplotlib 1.4.3 (both from condas repository), Win7

[minrk]

I think matplotlib requires PIL for jpg export to work:

conda install pillow

[jankatins]

Shouldn't this then be:

supported = {'png', 'png2x', 'retina', 'svg', 'pdf'}
try:
    from PIL import Image
    supported = supported.union({'jpg', 'jpeg'})
except:
    pass 

[Carreau]

Probably not, cause this would import PIL which is relatively heavy, even if people don't use it.

[minrk]

And the export formats only requires PIL for certain backends (Agg/inline, but not GTK or OS X).

[Carreau]

Hi,

This issue has been quiet for almost 2 years, and I don't think there is anything we can do. It has been marked as no-action, and I'm going to close it to attempt to reduce the number of opened issue. Feel free to reopen if needed.

Thanks.

[imahgrefteh]

pip3 install pillow got it to work for me

[matrix-1992]

pip install pillow also can't solve the problem in my computer.....

[jash101]

pip install pillow also can't solve the problem in my computer.....

Try uninstalling pillow first and then install it back. Worked for me.
sudo pip3 uninstall pillow
sudo pip3 install pillow

[berkayyildirim]

I had encountered the same problem and solutions above had not been able to solve my issue. After that, I just tried plt.savefig("images/pie.png") by replacing .jpg with .png in .py file as well as in my .html file. Finally, pie.png was automatically added to my folders and my problem was solved.

[tacaswell]

This is due to a bug in Matplotlib (see matplotlib/matplotlib#16083) in how we were checking the version of pillow which was removed in pillow 7.0. This has been fixed on the relevant branches upstream and bug-fix releases should go out soon. For now, downgrade pillow.

[pablobots]

pip install "pillow<7" got it to work for me