once #4 is merged, we can refactor to draw bundles with a new BundleWidget

using chrome headless via docker, we can create test notebooks for the package.

Might be good to be able to export both static and interactive visualizations from this, even in a headless mode, and without running the notebook.

Typing:

test = NiftiWidget('/coefs/wholebrain30.nii') test.nifti_plotter(plotting_func=plotting.plot_glass_brain, threshold=(0.0, 10.0, 0.01), display_mode=['ortho','xz'])

Causes this error:

FileNotFoundError: [Errno 2] No such file or directory: '/coefs'

If I specify the path alternatively as './coefs/wholebrain30.nii', it complains:

TypeError: argument should be string, bytes or integer, not PosixPath

e.g.

The current masking technique doesn't work well.

It should be improved by:

• checking not too much of the image is masked
• allowing standard masks by e.g. passing 'mne'

Any idea what might be going on? The volume viewer works great.

The sagital view is rendering with the eyes facing right however convention is eyes facing left. (e.g. see here).

Related to #20

In the axial view, the right side of the brain should render on the left side of the image, but instead it renders on the right.

See for example: https://www.dropbox.com/s/glaioa61bjj12lc/IBSR_01_ana.nii?dl=0 where the right lateral ventricle is mask value 43 (left is 4; see: https://www.nitrc.org/frs/shownotes.php?release_id=2316).

Hi,

I'm having a lot of trouble with this package on Google Colab:

!pip install --quiet niwidgets
from niwidgets import NiftiWidget

---------------------------------------------------------------------------
ContextualVersionConflict                 Traceback (most recent call last)
<ipython-input-2-80347fcf605d> in <module>()
      1 get_ipython().system('pip install --quiet niwidgets')
----> 2 from niwidgets import NiftiWidget

5 frames
/usr/local/lib/python3.6/dist-packages/pkg_resources/__init__.py in resolve(self, requirements, env, installer, replace_conflicting, extras)
    789                 # Oops, the "best" so far conflicts with a dependency
    790                 dependent_req = required_by[req]
--> 791                 raise VersionConflict(dist, req).with_context(dependent_req)
    792 
    793             # push the new requirements onto the stack

ContextualVersionConflict: (nibabel 2.3.3 (/usr/local/lib/python3.6/dist-packages), Requirement.parse('nibabel<3.0,>=2.4'), {'niwidgets'})

Can the requirements be made a bit more flexible or updated?

Here's the notebook: https://colab.research.google.com/drive/1k6QK75t0Ab1Zo-_j2YG3x85EvRSVjGLT

Thanks

As I see this (I might be wrong), when specifying the plotting_func the widget creates internally a new figure in any case, and passing then a custom figure handle as argument creates an error.

from niwidgets import NiftiWidget
import nilearn.plotting as nip

my_widget = NiftiWidget(img)
# my_widget.nifti_plotter(figsize=(10,20))

fig, axes = plt.subplots(1, 3, figsize=(5, 10))
my_widget.nifti_plotter(plotting_func=nip.plot_img, colormap=None, figsize=(5, 10),
                        figure=fig,
                        axes = axes)

The default case works fine, the img is shown as 3 axes and the figsize can be adjusted. To get more control over the figure and axes, I used the function call shown above but get the follow error message:

---------------------------------------------------------------------------
ValueError                                Traceback (most recent call last)
<ipython-input-92-a281499c2288> in <module>()
     33 my_widget.nifti_plotter(plotting_func=nip.plot_img, colormap=None, figsize=(5, 10),
     34                         figure=fig,
---> 35                         axes = axes
     36                        )

~/anaconda3/lib/python3.6/site-packages/niwidgets/niwidget_volume.py in nifti_plotter(self, plotting_func, colormap, figsize, **kwargs)
     83             self._default_plotter(**kwargs)
     84         else:
---> 85             self._custom_plotter(plotting_func, **kwargs)
     86 
     87     def _default_plotter(self, mask_background=False, **kwargs):

~/anaconda3/lib/python3.6/site-packages/niwidgets/niwidget_volume.py in _custom_plotter(self, plotting_func, **kwargs)
    241 
    242         # Create the widget:
--> 243         interact(self._custom_plot_wrapper, data=fixed(self.data), **kwargs)
    244         plt.ion()
    245 

~/anaconda3/lib/python3.6/site-packages/ipywidgets/widgets/interaction.py in __call__(self, _InteractFactory__interact_f, **kwargs)
    521         #    def f(*args, **kwargs):
    522         #        ...
--> 523         w = self.widget(f)
    524         try:
    525             f.widget = w

~/anaconda3/lib/python3.6/site-packages/ipywidgets/widgets/interaction.py in widget(self, f)
    437             The function to which the interactive widgets are tied.
    438         """
--> 439         return self.cls(f, self.opts, **self.kwargs)
    440 
    441     def __call__(self, __interact_f=None, **kwargs):

~/anaconda3/lib/python3.6/site-packages/ipywidgets/widgets/interaction.py in __init__(self, _interactive__interact_f, _interactive__options, **kwargs)
    193             getcallargs(f, **{n:v for n,v,_ in new_kwargs})
    194         # Now build the widgets from the abbreviations.
--> 195         self.kwargs_widgets = self.widgets_from_abbreviations(new_kwargs)
    196 
    197         # This has to be done as an assignment, not using self.children.append,

~/anaconda3/lib/python3.6/site-packages/ipywidgets/widgets/interaction.py in widgets_from_abbreviations(self, seq)
    293             if not (isinstance(widget, ValueWidget) or isinstance(widget, fixed)):
    294                 if widget is None:
--> 295                     raise ValueError("{!r} cannot be transformed to a widget".format(abbrev))
    296                 else:
    297                     raise TypeError("{!r} is not a ValueWidget".format(widget))

ValueError: <Figure size 360x720 with 0 Axes> cannot be transformed to a widget

<Figure size 360x720 with 0 Axes>

When passing only the axes handles, the error is

AssertionError: The axes passed are not in the figure

I am getting an error related to specifying a path.

TypeError: argument should be string, bytes or integer, not PosixPath

This was with the pip install niwidgets basic installation; I had trouble with this on 2.7.2 so I switched my kernel to 3.5.3 and am still having the same issue...

TypeError Traceback (most recent call last)
in ()
5 import ipywidgets as widgets
6 ###http://nipy.org/niwidgets/installation.html#install-via-pip
----> 7 my_widget = NiftiWidget('../SampleDataFiles/fdt_paths.nii.gz')
8 get_ipython().run_line_magic('load_ext', 'autoreload')
9 get_ipython().run_line_magic('autoreload', '2')

~/.virtualenvs/nipypev3/lib/python3.5/site-packages/niwidgets/niwidget_volume.py in init(self, filename)
43 """
44 self.filename = Path(filename).resolve()
---> 45 if not os.path.isfile(self.filename):
46 raise OSError('File ' + self.filename.name + ' not found.')
47

~/.virtualenvs/nipypev3/lib/python3.5/genericpath.py in isfile(path)
28 """Test whether a path is a regular file"""
29 try:
---> 30 st = os.stat(path)
31 except OSError:
32 return False

TypeError: argument should be string, bytes or integer, not PosixPath

I'm currently trying to use the streamline widget to plot some output, however I've noticed that the visualization always starts extremely far away (to the point where a user might thing there's nothing actually plotted).

I did some exploring (to the extent that I can) and it seems that the visualization is defaulting to a suspiciously arbitrary value:

catchPlot=sw.plot(display_fraction=1, width=1000, height=1000, style=style, percentile=0) VBox(children=(Figure(camera=PerspectiveCamera(fov=46.0, position=(0.0, 0.0, 2.0), quaternion=(0.0, 0.0, 0.0, … interactive(children=(FloatSlider(value=12.84813404083252, continuous_update=False, description='threshold', m…

This value appears to be the default FOV (as well as the other settings) as found in the ipyvolume documentation

Additionally, another ipyvolume user noted a similar behavior previously.

Is this a ipyvolume issue, a niwidget issue, or am I just using niwidgets incorrectly?

This project lists jupyter-sphinx in its sphinx configuration or requirements file, but does not appear to use this extension directly anymore.
We have just rewritten jupyter-sphinx from scratch, and fully changed the interface.
Therefore we would like to give you a heads up, in case you find the new version useful.

Now all code execution is done via a single jupyter-execute sphinx directive, instead of the previous ipywidgets-setup/ipywidgets-embed.
The new version is also much more flexible: it allows to embed any output recognized by jupyter in sphinx documentation. See the project documentation for details.

We have released 0.2.0rc1 on pip, please give it a shot using pip install 0.2.0rc1 --pre — we would love to hear your feedback.
Please also let us know by responding here or in jupyter/jupyter-sphinx#33 if the new release disrupts your workflow.

Just immediately after installation I get the following errror on Python 3.6.5:

from niwidgets import NiftiWidget

---------------------------------------------------------------------------
ImportError                               Traceback (most recent call last)
<ipython-input-1-c8f24c578dcb> in <module>()
----> 1 from niwidgets import NiftiWidget

/home/arman/anaconda/envs/development/lib/python3.6/site-packages/niwidgets/__init__.py in <module>()
      8 from .exampledata import exampleatlas, examplezmap, examplet1  # noqa: F401
      9 from .niwidget_volume import NiftiWidget  # noqa: F401
---> 10 from .niwidget_surface import SurfaceWidget  # noqa: F401
     11 from .streamlines import StreamlineWidget  # noqa: F401

/home/arman/anaconda/envs/development/lib/python3.6/site-packages/niwidgets/niwidget_surface.py in <module>()
      6 from xml.parsers.expat import ExpatError
      7 
----> 8 import ipyvolume.pylab as p3
      9 import matplotlib.pyplot as plt
     10 import nibabel as nb

/home/arman/anaconda/envs/development/lib/python3.6/site-packages/ipyvolume/__init__.py in <module>()
      3 
      4 from . import styles
----> 5 from .widgets import *
      6 from .transferfunction import *
      7 from . import examples

/home/arman/anaconda/envs/development/lib/python3.6/site-packages/ipyvolume/widgets.py in <module>()
      8 import logging
      9 import numpy as np
---> 10 from .serialize import array_cube_tile_serialization, array_serialization, array_sequence_serialization,\
     11     color_serialization, image_serialization, texture_serialization
     12 from .transferfunction import *

/home/arman/anaconda/envs/development/lib/python3.6/site-packages/ipyvolume/serialize.py in <module>()
      7 from . import utils
      8 import ipywidgets
----> 9 import ipywebrtc
     10 import numpy as np
     11 import PIL.Image

/home/arman/anaconda/envs/development/lib/python3.6/site-packages/ipywebrtc/__init__.py in <module>()
      3 import ipywidgets as widgets
      4 from ._version import version_info, __version__
----> 5 from .webrtc import *
      6 
      7 

/home/arman/anaconda/envs/development/lib/python3.6/site-packages/ipywebrtc/webrtc.py in <module>()
     10     Bool, Bytes, Dict, Instance, Int, List, TraitError, Unicode, validate
     11 )
---> 12 from ipywidgets import DOMWidget, Image, Video, Audio, register, widget_serialization
     13 from ipython_genutils.py3compat import string_types
     14 import ipywebrtc._version

ImportError: cannot import name 'Video'

I just tried to pip install from a directory I cloned from github and when I import NiftiWidget, I get the following error:

---------------------------------------------------------------------------
TypeError                                 Traceback (most recent call last)
<ipython-input-1-ab0954c2e00f> in <module>()
      7 import seaborn as sns
      8 import numpy as np
----> 9 from niwidgets import NiftiWidget
     10 
     11 get_ipython().magic('matplotlib inline')

/data/MLcore/miniconda3/envs/nielsond/lib/python3.6/site-packages/niwidgets/__init__.py in <module>()
      9 # import widget classes.
     10 from .niwidget_volume import NiftiWidget  # noqa
---> 11 from .niwidget_surface import SurfaceWidget  # noqa

/data/MLcore/miniconda3/envs/nielsond/lib/python3.6/site-packages/niwidgets/niwidget_surface.py in <module>()
      7 from ipywidgets import interact, fixed, IntSlider
      8 # from ipyvolume import gcf
----> 9 import ipyvolume.pylab as p3
     10 import os
     11 # import pathlib & backwards compatibility

/data/MLcore/miniconda3/envs/nielsond/lib/python3.6/site-packages/ipyvolume/__init__.py in <module>()
      3 
      4 from . import styles
----> 5 from .widgets import *
      6 from .transferfunction import *
      7 from .examples import  *

/data/MLcore/miniconda3/envs/nielsond/lib/python3.6/site-packages/ipyvolume/widgets.py in <module>()
     10 from .serialize import array_cube_png_serialization, array_serialization, array_sequence_serialization,\
     11     color_serialization, image_serialization, texture_serialization
---> 12 from .transferfunction import *
     13 import warnings
     14 import ipyvolume

/data/MLcore/miniconda3/envs/nielsond/lib/python3.6/site-packages/ipyvolume/transferfunction.py in <module>()
     38 
     39 
---> 40 class TransferFunctionJsBumps(TransferFunction):
     41         _model_name = Unicode('TransferFunctionJsBumpsModel').tag(sync=True)
     42         _model_module = Unicode('ipyvolume').tag(sync=True)

TypeError: function() argument 1 must be code, not str

Related to #20

I think the brain stem should be at the top (https://www.intechopen.com/source/html/38664/media/image1.jpeg) but is rendering on the bottom.

Terrific widgets package you made, great job!

I tried the online interactive demo, but none of the notebooks work, probably the package is not updated there?

Also the link to mybinder is buggy in the documentation:
http://nipy.org/niwidgets/examples.html

xposted: jupyter-widgets/ipywidgets#1904

Goes away after moving slider:

I am trying to plot a volume of float32 data (takes on values 0-1) and I get the following:

If instead I convert the float data to uint8 (*255), the plotting looks fine (not totally washed out).

Likewise If I just do:

plt.imshow(img_float.dataobj[...,87], vmin=0., vmax=1., cmap='viridis')

I also am able to see the image. What is happening in niwidgets to wash out the image?

Right now only python 3.6 is supported:

/usr/src/app/src/niwidgets/niwidgets/niwidget_volume.py in __init__(self, filename)
     41                     ``PosixPath`` from python3's pathlib.
     42         """
---> 43         self.filename = Path(filename).resolve(strict=True)
     44 
     45         # load data in advance

TypeError: resolve() got an unexpected keyword argument 'strict'

due to strict being added in 3.6.

NiftiWidget should also allowed to be constructing providing the nibabel image. Right now only a path is allowed. load is this called on the path. I suggest allowing the user to pass in the same (ie the result of the user having called load himself) into NiftiWidget.

Alternatively, NiftiWidget can be made to subclass VolumetricWidget (handles more than nii files) and then VolumetricWidget takes in the nibabel image.

Clarify what version(s) of Python are supported.

• using python_requires in setup.py
• in the README

Right now it looks like testing is agaisnt python 3.x but that does not seem to be made explicit.

Ahoi hoi there,

Problem:

I was just trying to run the examples on mybinder, but appears to be a problem as none of the imports do work:

Steps to reproduce:

Try to run the examples on mybinder here.

Possible solution:

No precise idea, but seems that the examples are running on a local mybinder instance. Hence, maybe check if everything is okay there.

Happy to help solving this in case there's something I could do.

Best, Peer

Hi

i use this package for simple cardiac stack visualization. i'm writing a 3d analysis script for orientation and volumetric estimation of cardiac cells.

The issue is the rotation and the direction of the axes.

I explain it with an example:

I have a stack with M images if NxN pixels. If i plot one slice with Imagej (or matplotlib.imshow) i see correct orientation, when i create a Nibabel image with
my_stack_nib = nib.Nifti1Image(my_stack, affine) # with affine = numpy.eye(4) from your example

and plot it with
widget = niwidgets.NiftiWidget(my_stack_nib)
widget.nifti_plotter()

i see :

• the xy plane rotated by 90' clockwise.
• the z and y axex inverted respect to numpy stack information ( z = 0 on bottom and z = zmax in up, y = 0 in bottom and y = ymax in up).

To fix it , i tried to change
nifti_plotter(plotting_func=...)

i tried numpy.rotate on xy plan (OK) and flip on z axe (but this breaks correct reference from data to plot)
i tried to change affine matrix, but it seems not to works.

I search only a simple way to change axes direction without corrupt my data, please someone can help me?

thanks a lot
Dr. Francesco Giardini, LENS, Florence

@janfreyberg and @arokem - i think we should change the license to apache 2. cc0 doesn't really apply well to software.

The gif presented in the readme.md file depicting volumetric image plotting features a "value at point" output behavior that appears to print out the value at the specified voxel.

However, this does not appear to be a default behavior of the current version of the widget. Furthermore, a GitHub search of the repo for "value at point" does not return any hits. Is there a hidden option for this behavior? Has this functionality been removed?

Ahoi hoi there,

Problem:

Trying to get to the documentation through the link in the README leads to a 404, as the link appears to be broken:

The link is https://github.com/nipy/niwidgets/blob/master/nipy.org/niwidgets.

Steps to reproduce:

Clicking on the documentation link in the README.

Possible solution:

Fix the link to lead to the working documentation, which appears to be
https://nipy.org/niwidgets/

Happy to submit a short PR for this in case you want.

Best, Peer

it would be nice to update the documents to reflect the new widgets. if we use nbconvert and embed widgets we could also be able to generate interactive docs to demonstrate. see the ipywidgets docs for an example.

import niwidgets

---------------------------------------------------------------------------
AttributeError                            Traceback (most recent call last)
<ipython-input-2-c2cf98e8d4d7> in <module>()
----> 1 import niwidgets
      2 #from niwidgets import NiftiWidget

D:\Anaconda3\lib\site-packages\niwidgets\__init__.py in <module>()
      9 # import widget classes.
     10 from .niwidget_volume import NiftiWidget  # noqa
---> 11 from .niwidget_surface import SurfaceWidget  # noqa

D:\Anaconda3\lib\site-packages\niwidgets\niwidget_surface.py in <module>()
      7 from ipywidgets import interact, fixed, IntSlider
      8 # from ipyvolume import gcf
----> 9 import ipyvolume.pylab as p3
     10 import os
     11 # import pathlib & backwards compatibility

D:\Anaconda3\lib\site-packages\ipyvolume\__init__.py in <module>()
      3 
      4 from . import styles
----> 5 from .widgets import *
      6 from .transferfunction import *
      7 from .examples import  *

D:\Anaconda3\lib\site-packages\ipyvolume\widgets.py in <module>()
     10 from .serialize import array_cube_png_serialization, array_serialization, array_sequence_serialization,\
     11     color_serialization, image_serialization, texture_serialization
---> 12 from .transferfunction import *
     13 import warnings
     14 import ipyvolume

D:\Anaconda3\lib\site-packages\ipyvolume\transferfunction.py in <module>()
     11 
     12 import ipyvolume._version
---> 13 semver_range_frontend = "~" + ipyvolume._version.__version_js__
     14 
     15 

AttributeError: module 'ipyvolume' has no attribute '_version'

And i think the question is about jupyter extensions cause that i've tried the code in anaconda prompt and failed:
jupyter nbextension enable --py --sys-prefix ipyvolume

It would be great to change the style of the docs to the readthedocs style - the current style isn't very discoverable.

Hey there,

I installed niwidgets from pypi and when doing this:

from niwidgets import NiftiWidget

I get the following error:

---------------------------------------------------------------------------
NameError                                 Traceback (most recent call last)
<ipython-input-3-580c792c4ebb> in <module>()
      8 from nilearn import image
      9 from nilearn import plotting
---> 10 from niwidgets import NiftiWidget
     11 
     12 try:

C:\Anaconda2\lib\site-packages\niwidgets\__init__.py in <module>()
      6 """
      7 # import example data
----> 8 from .exampledata import exampleatlas, examplezmap, examplet1  # noqa
      9 # import widget classes.
     10 from .niwidget_volume import NiftiWidget  # noqa

C:\Anaconda2\lib\site-packages\niwidgets\exampledata.py in <module>()
      4     # on >3 this ships by default
      5     from pathlib import Path
----> 6 except ModuleNotFoundError:
      7     # on 2.7 this should work
      8     try:

NameError: name 'ModuleNotFoundError' is not defined

I am using Python v2.7 BTW

Hi niwidgets developers, currently I am only able to get your package via a pip wheel. The instructions in the README to run

pip install git+git://github.com/nipy/niwidgets/

doesn't work due to a missing setup.py. Any advice on how to build from git/source?

Hi all,

I get the following FutureWarning from numpy when calling nifti_plotter:

It seems to be caused by line 197 in niwidget_volume.py:

else np.fliplr(np.flipud(np.rot90(data[slice_obj], k=1)))

I think it doesn't have any real effect other than displaying the annoying warning. However, it could be nice to either:

Option A

Add:

import warnings
warnings.simplefilter(action='ignore', category=FutureWarning)

before importing numpy.

Option B

Even simpler, change line 197 to:

else np.fliplr(np.flipud(np.rot90(data[tuple(slice_obj)], k=1)))

I tested both solutions locally and they seem to work as expected.

If that makes sense to you I will gladly create a PR.

Thank you for all your amazing work!