New function to accept multiple images for segmentation #91

JohnnyTeutonic · 2017-06-01T07:47:44Z

I have included a new function to accept multiple images in viz. This function accepts multiple images in its function call to allow for a convenient display of various segmented images in one figure. It also includes an optional parameter which allows for the displaying of the images in different colormaps.
I have also included a cast to int inside of imshow_rand when the random colors are being displayed, as I noticed that there was an issue with calling this function (owing to the use of np.ceil).

coveralls · 2017-06-01T07:54:04Z

Coverage decreased (-0.1%) to 50.472% when pulling 4943a84 on JohnnyTeutonic:imshow into f84a682 on janelia-flyem:master.

coveralls · 2017-06-02T03:35:46Z

Coverage decreased (-0.2%) to 50.459% when pulling 88f02b4 on JohnnyTeutonic:imshow into f84a682 on janelia-flyem:master.

coveralls · 2017-06-02T06:10:12Z

Coverage decreased (-0.2%) to 50.406% when pulling 574d00e on JohnnyTeutonic:imshow into f84a682 on janelia-flyem:master.

coveralls · 2017-06-02T06:52:22Z

Coverage decreased (-0.2%) to 50.392% when pulling 8a4009f on JohnnyTeutonic:imshow into f84a682 on janelia-flyem:master.

jni

@JohnnyTeutonic have you attempted to track down the LAB range issues from imshow_rand? Otherwise, I've added some minor comments.

jni · 2017-06-02T07:44:28Z

gala/viz.py

@@ -55,20 +55,61 @@ def imshow_rand(im, labrandom=True):
    fig : plt.Figure
        The image shown.
    """
-    rand_colors = np.random.rand(np.ceil(im.max()), 3)
+    rand_colors = np.random.rand(int(np.ceil(im.max())), 3)


Let's replace this with:

from math import ceil # <-- near top of file [...] rand_colors = np.random.random(size=(ceil(np.max(im)), 3))

jni · 2017-06-02T07:55:27Z

gala/viz.py

    return plt.imshow(im, cmap=rcmap, interpolation='nearest')


+def multiple_images(*images, raw=False, image_type='rand'):


What's the use case here? Show two segmentations of the same image side by side?

I actually like the basic idea for this function, but the current interface hampers its usability. For example, it might be good to display the original image as well, but that would require a different colormap...

Also it needs a "show" in the name.

Hey, I've already incorporated the ability to show the original image side-by-side with the other images. The use case is to allow for the inclusion of multiple segmented images alongside an original image. This is going to back to a conversation we had a few weeks ago. I have included the parameter 'raw' which lets you get the raw versions of the images. If you want to have a randomised color image alongside the raw image, then it will do this by default for you.

But the colormap is the same for every displayed image, right? Whereas normally you would want grey/magma for the raw image, and randomised colours for the segmentation?

Well yeah my code is a bit hacky right now but I know what you are asking for. I have been trying to incorporate that suggestion today and hadn't had much luck. I'm sure I'll figure out how to do that pretty soon though as it would be neat to be able to do that.

jni · 2017-06-02T07:56:25Z

gala/viz.py

+    for i in range(1, number_of_im+1):
+        ax = figure.add_subplot(1, number_of_im, i)
+        if image_type == 'grey' or image_type == 'gray':
+            imshow_grey(images[i-1])


Add an axes= kwarg to the other imshow functions so that those specific axes are used for the imshow.

I just noticed this: maybe iterate in range(number_of_im) and then you just need one i+1 instead of several i-1s?

Not sure what you mean here?

for i in range(1, number_of_im+1): ... images[i-1]

When I said 'I'm not sure what you mean', I was referring to what I didn't understand what you meant by adding the axes kwarg.

Yeah, the looping part started like that because I mistakenly believed that a call to figure.add_subplot required indexing beginning from one, but I have changed that now...

I have just committed some changes that include the axis parameter but am not sure about the implementation so will wait to hear from you I guess

jni · 2017-06-02T07:56:54Z

gala/viz.py

+        if image_type == 'grey' or image_type == 'gray':
+            imshow_grey(images[i-1])
+        elif image_type == 'jet':
+            imshow_jet(images[i-1])


I think we can delete Jet altogether, as it is an outdated colormap. My current preference would be imshow_magma to replace it. =)

jni · 2017-06-02T07:57:53Z

gala/viz.py

+        ax.set_title("Image no. {} with a ''{}'' colormap.".
+                     format(i, image_type))
+        if raw:
+            plt.imshow(images[i-1])


Use ax.imshow instead of plt.imshow. See Elegant SciPy for some examples (this is the interface that MPL is moving towards, and it makes more sense to use specific objects, rather than rely on some global state).

I get your reasoning there :)

jni · 2017-06-02T09:20:01Z

gala/viz.py

+                     format(i, image_type))
+        if raw:
+            plt.imshow(images[i-1])
+            ax.set_title("Image no. ''{}'' .".format(i))


Let's use Python 3.6 syntax: ax.set_title(f'Image number {i}')

JohnnyTeutonic · 2017-06-02T13:49:57Z

Also, got big-cat launch to work on linux!

JohnnyTeutonic · 2017-06-02T14:44:03Z

I think the build is failing because it is building with python3.5 and I have included the 3.6 formatting syntax as you have suggested.
Also, I missed your comment about the LAB coloring scheme. I have investigated that and can tell you that the warning is caused by the 'z' axis in the call 'skimage/color/colorconv.py/lab2xyz' where there are pixels with a value lower than zero. I haven't been figure out how to change the code to deal with this warning as yet (barring the explicit suppression of that particular warning) but I can tell you that this doesn't occur when the parameter 'labrandom' is set to false in imshow_rand.

JohnnyTeutonic · 2017-06-03T01:03:05Z

I had to change the looping back to 'i-1' because plt.subplot has indexing start at 1 for compatibility purposes with MATLAB :)

jni · 2017-06-03T03:02:53Z

@JohnnyTeutonic you can (and should) change the Python version used for testing in .travis.yml.

Re i+1, I know that subplots use that, I'm saying, make i start from 0 and then use +1 where you have to, rather than making it start from 1 and then using -1 where you have to. =)

Regarding passing axes, here's what I mean:
https://matplotlib.org/faq/usage_faq.html#coding-styles

ie instead of:

plt.imshow(image)

do:

if ax is None:
    fig, ax = plt.subplots()
ax.imshow(image)

or:

if ax is None:
    ax = plt.gca()  # get current axes
ax.imshow(image)

Generally, always use ax.imshow and ax.plot, instead of plt.. I know, I know, the existing code doesn't do that. I was young and foolish when I wrote it. Much like you are now. =P =P

jni · 2017-06-03T03:06:46Z

@JohnnyTeutonic regarding the lab2xyz -> xyz2rgb color conversion, I know (vaguely) that that's what's causing it, what I'm saying is, I would like you to figure out when a Lab triplet is out of range and fix it before converting to RGB. =) OR, you could manually convert lab2xyz, use np.clip to clip the negative z values, and then use xyz2rgb to get the RGB tuples back.

I want to use labrandom basically always because, to my eyes at least, it produces a much more pleasing color map than random RGB. Don't you think?

JohnnyTeutonic · 2017-06-03T03:52:13Z

Well I tried changing it to 3.6 in the .yml file but still failed. I have tried changing it in setup.py as well to see if that helps...
I changed all instances to ax.imshow.
I thought of using np.clip last night actually on that! I wasn't able to succeed .I also tried manually changing the values but still have a few pixels get below zero. I will hopefully solve this issue soon :)
And yes, I am young and foolish that is true!

JohnnyTeutonic · 2017-06-03T05:46:00Z

No module named cythonize???

JohnnyTeutonic · 2017-06-03T05:46:22Z

sorry, no module named Cython...Very confused.

JohnnyTeutonic · 2017-06-03T05:50:18Z

I am wondering whether Cython 0.17 is available yet for python 3.6?

JohnnyTeutonic · 2017-06-03T06:16:38Z

Well I got rid of the cython error but I am still not succeeding in getting the test to pass.

JohnnyTeutonic · 2017-06-03T06:29:39Z

I have changed the setup.cfg file, asv.conf.json and setup.py to reflect the adoption of 3.6. I am unsure whether this is what I should have done but obviously I can rollback if need be. Looking at the Travis build, I can see that the virtual env is being created in 3.6 but the tests are being run still in 3.5. I have no idea why...

JohnnyTeutonic · 2017-06-03T08:54:44Z

I've tried for several hours now to solve the problem with the Z pixels. I have tried clipping as well as altering the values for 'rand_colors' to no avail. The issue to me seems to be with the code inside skimage/color/colorconv.py, Even if you perform np.clip before calling lab2xyz or lab2rgb, it will still complain about Z pixels being below 0. It's not very straightforward to determine values that will not cause Z pixels to be below 0.

JohnnyTeutonic · 2017-06-03T09:30:15Z

Okay, sorry for all the comments. I tweaked the colours now and the warning is gone...

jni

@JohnnyTeutonic I think we'll just need one last round to get this in! =)

Don't worry about Travis for now. Your ongoing problems had to do with Travis caching the conda environment (see the "cache" statement at the end of the yaml file). I'll have to explore ways for you to clear it during PRs (though I hope we don't have to do so again any time soon).

The current build failure is because my tree merging package viridis hasn't been uploaded for Python 3.6 for conda. I'll update that shortly.

Thanks again and sorry for the Travis confusion!

jni · 2017-06-04T00:58:45Z

gala/evaluate.py

    ii1 = back1[i1] if compress else i1
    ii2 = back2[i2] if compress else i2
-    return ii1, h2g1[i1], ii2, h1g2[i2]
+    return ii1, h1g2[i1], ii2, h2g1[i2]


Nice! Just as a warning, I'm gonna get you to explain this change to me on Tuesday, cos it's still murky in my own head. =)

Happy to do so!

jni · 2017-06-04T00:59:05Z

gala/viz.py


 ###########################
 # VISUALIZATION FUNCTIONS #
 ###########################

-def imshow_grey(im):
+def imshow_grey(im, axis=''):


Here, axis should default to None.

jni · 2017-06-04T01:02:32Z

gala/viz.py

@@ -21,11 +22,11 @@ def imshow_grey(im):
    fig : plt.Figure
        The image shown.
    """
-    return plt.imshow(im, cmap=plt.cm.gray, interpolation='nearest')
+    return plt.imshow(im, cmap='gray', interpolation='nearest')


And then here, you should have:

if axis is None: fig, axis = plt.subplots() return axis.imshow(im, cmap='gray')

Incidentally, interpolation= is no longer required, since it's the new default in MPL 2.0:
https://matplotlib.org/users/dflt_style_changes.html#interpolation

Also, plt.subplots makes a new figure. You could also use plt.gca (get current axes) to plot on the currently active figure by default. I'm ambivalent about which should be the default.

jni · 2017-06-04T01:03:24Z

gala/viz.py



-def imshow_jet(im):
-    """Show a segmentation using a jet colormap.
+def imshow_magma(im, axis=''):


Same thing here.

jni · 2017-06-04T01:04:15Z

gala/viz.py

-        rand_colors[:, 2] = rand_colors[:, 2] * 198 - 106
+        rand_colors[:, 0] = rand_colors[:, 0] * 81 + 39
+        rand_colors[:, 1] = rand_colors[:, 1] * 185 - 86
+        rand_colors[:, 2] = rand_colors[:, 2] * 198 - 108


Can you document where you got these values from? (ie add a comment inside the block with a link) But, thanks for fixing it!!! =) Been living with these errors for years. =D

jni · 2017-06-04T01:06:48Z

gala/viz.py

    return plt.imshow(im, cmap=rcmap, interpolation='nearest')


+def show_multiple_images(*images, ax='', raw=False, image_type='rand'):


Here, I would have axes=None, and be able to optionally pass in a list of axes that should be of the same length as images. Then each image would be plotted on the corresponding axis.

jni · 2017-06-04T01:08:29Z

gala/viz.py

+            try:
+                imshow_rand(images[i])
+            except ValueError:
+                ax.imshow(images[i], axis=ax)


This just needs ax.imshow, the axis will then be correctly selected. (indeed I'm not sure but this might just fail as currently written)

Originally it failed...I will see what happens with the new code though!

jni · 2017-06-04T01:10:19Z

gala/viz.py

+    number_of_im = len(images)
+    figure = plt.figure()
+    for i in range(number_of_im):
+        ax = figure.add_subplot(1, number_of_im, i+1)


Then, here, you can say:

ax = (figure.add_subplot(1, number_of_im, i+1) if axes is None else axes[i])

jni · 2017-06-04T02:50:56Z

WTF:
https://travis-ci.org/janelia-flyem/gala/builds/239038966#L495

Probably some update to networkx that broke the edge iteration. Can you update the test to:

assert (1, 5) in g.edges() or (5, 1) in g.edges()

Not due to your PR but would be nice to have the tests passing before merging. =)

JohnnyTeutonic · 2017-06-04T03:32:06Z

I have tried to include the changes you suggested. I am a bit confused as to how to pass in the 'axes' parameter in my own function show_multiple_images. My use case is to allow for just one call of my function to show multiple images in the same figure. I have included your changes but am not sure how to pass in the axes object inside my function. I have looked at some documentation but still am a bit confused.
Also, I fixed the LAB conversion scheme through randomising the input values. I didn't find a formula...It still is randomised however. I can't see any segments that have the same colour lying contiguously.

coveralls · 2017-06-04T03:33:18Z

Coverage decreased (-0.3%) to 50.327% when pulling 981755e on JohnnyTeutonic:imshow into f84a682 on janelia-flyem:master.

jni · 2017-06-04T05:39:01Z

@JohnnyTeutonic the idea is that you can create the figure ahead of time:

images = ...
n_images = len(images)
fig, axes = plt.subplots(1, len(images))
show_multiple_images(images, axes, 'rand')

(Incidentally, if axes is not None, you probably want to remove the call to plt.figure().)

That way, you now have the flexibility to use some other grid style:

images = ...  # 4 images
fig, axes = plt.subplots(2, 2)
show_multiple_images(images, axes=axes.ravel(), 'rand')

And, this reminded me of something. Rather than using plt.figure and fig.add_axes, you should probably use plt.subplots at the top — this will let you use sharex=True, sharey=True, which will give you joint panning and zooming for all of the panels in your figure.

coveralls · 2017-06-04T05:41:36Z

Coverage decreased (-0.3%) to 50.327% when pulling b221672 on JohnnyTeutonic:imshow into f84a682 on janelia-flyem:master.

JohnnyTeutonic · 2017-06-04T06:27:44Z

I will see if i can wrap my head around it. Incidentally, ev.sorted_vi_components I think has some buggy code....I will try and show you it in person on Tuesday. It is not giving me the desired labels that I need.
Here is the output of my notebook showing what I mean with a simple segmentation. Please forgive me if I have mistaken something conceptually. When I call sorted_vi_components, I believe it should be outputting just one label corresponding to a label which has the worst entropy (h(y|x)) found in the gt and one label that corresponds to the worst entropy for h(x|y) for the input segment. If you look at my notebook, you will see that it recursively outputs four four-tuples (giving 16 values), when it should be only outputting a single four-tuple with the label in gt with worst entropy, h(y|x), the label in seg with worst entropy, h(x|y) or whatever the order is... (I fixed the docstring as you know).

You should see that some of the numbers are floating poitn numbers. They definitely correspond to the entropy scores. I think I can maybe fix the code up but only if you think I am right about this...

jni · 2017-06-04T07:17:56Z

It's not buggy, it's clearly (for once =P) documented in the docstrings:

gala/gala/evaluate.py

Line 1214 in f84a682

Returns
gala/gala/evaluate.py

Line 1258 in f84a682

comps : list of (int, float, float) tuples

jni · 2017-06-04T07:18:53Z

btw let's keep the discussion on here related to the PR.

Also, try to use gist.github.com to share notebooks, rather than screengrabs. You can upload a full notebook to a private gist and then share the URL.

JohnnyTeutonic · 2017-06-04T07:41:31Z

I read the docstrings of course but I found it a bit difficult to follow what was going on I guess, owing to the complexity of the format. Anyway, I get now that the first tuple is the labels, second is entropy, third is other labels, fourth is other entropy. all good.

jni · 2017-06-04T08:23:26Z

Here's the NumPy Docstring standard format, which you should be familiar with. =)

https://github.com/numpy/numpy/blob/master/doc/HOWTO_DOCUMENT.rst.txt

coveralls · 2017-06-04T09:36:31Z

Coverage decreased (-0.3%) to 50.353% when pulling 75094bd on JohnnyTeutonic:imshow into f84a682 on janelia-flyem:master.

JohnnyTeutonic added 6 commits May 6, 2017 15:45

Add plt.subplots into viz.rand

a9dc972

Add function for multiple images

43de960

Add imshow_rand to function

0876336

Include cast to int in imshow_rand

27633dc

Include function to return multiple images

008d4c2

Include docstring for multiple_images function

4943a84

Include setting title of axes, docstring changes

88f02b4

Add parameter to include raw images

574d00e

Include try catch block

8a4009f

jni requested changes Jun 2, 2017

View reviewed changes

jni reviewed Jun 2, 2017

View reviewed changes

JohnnyTeutonic added 4 commits June 2, 2017 20:02

Include requested changes

5c51f9d

add axis parameter

4df67e4

Include change to i1 and i2

096a589

fix typo

d5a0c64

fix other typos

a3ea070

Change 'i' indexing for subplot

5908958

Change indexing for axis title

4d7325a

JohnnyTeutonic added 2 commits June 3, 2017 13:32

Change looping again

01c99a8

Change test to 3.6

4b4ec45

JohnnyTeutonic added 2 commits June 3, 2017 15:41

change back??

69b320c

Change to 3.6 in env.yml

e8fcb70

JohnnyTeutonic added 3 commits June 3, 2017 15:51

Change json file

a7e7679

Change dependencies?

5b29531

Change v to 3.6 in env.yml

99e6d29

change colors to remove warning

d8aa183

jni reviewed Jun 4, 2017

View reviewed changes

JohnnyTeutonic added 2 commits June 4, 2017 13:25

Include changes requested

efdfbec

Change test_mask

981755e

Change docstring for sorted_vi

b221672

Random changes

75094bd

jni merged commit 8d983e6 into janelia-flyem:master Jun 7, 2017

		return plt.imshow(im, cmap=rcmap, interpolation='nearest')


		def multiple_images(*images, raw=False, image_type='rand'):

		return plt.imshow(im, cmap=rcmap, interpolation='nearest')


		def show_multiple_images(*images, ax='', raw=False, image_type='rand'):

New function to accept multiple images for segmentation #91

New function to accept multiple images for segmentation #91

Conversation

JohnnyTeutonic commented Jun 1, 2017

coveralls commented Jun 1, 2017

coveralls commented Jun 2, 2017

coveralls commented Jun 2, 2017

coveralls commented Jun 2, 2017

jni left a comment

Choose a reason for hiding this comment

Choose a reason for hiding this comment

Choose a reason for hiding this comment

Choose a reason for hiding this comment

Choose a reason for hiding this comment

Choose a reason for hiding this comment

Choose a reason for hiding this comment

Choose a reason for hiding this comment

Choose a reason for hiding this comment

Choose a reason for hiding this comment

Choose a reason for hiding this comment

Choose a reason for hiding this comment

Choose a reason for hiding this comment

Choose a reason for hiding this comment

Choose a reason for hiding this comment

Choose a reason for hiding this comment

Choose a reason for hiding this comment

Choose a reason for hiding this comment

Choose a reason for hiding this comment

JohnnyTeutonic commented Jun 2, 2017

JohnnyTeutonic commented Jun 2, 2017

JohnnyTeutonic commented Jun 3, 2017

jni commented Jun 3, 2017

jni commented Jun 3, 2017

JohnnyTeutonic commented Jun 3, 2017

JohnnyTeutonic commented Jun 3, 2017

JohnnyTeutonic commented Jun 3, 2017

JohnnyTeutonic commented Jun 3, 2017

JohnnyTeutonic commented Jun 3, 2017

JohnnyTeutonic commented Jun 3, 2017

JohnnyTeutonic commented Jun 3, 2017

JohnnyTeutonic commented Jun 3, 2017

jni left a comment

Choose a reason for hiding this comment

Choose a reason for hiding this comment

Choose a reason for hiding this comment

Choose a reason for hiding this comment

Choose a reason for hiding this comment

Choose a reason for hiding this comment

Choose a reason for hiding this comment

Choose a reason for hiding this comment

Choose a reason for hiding this comment

Choose a reason for hiding this comment

Choose a reason for hiding this comment

jni commented Jun 4, 2017

JohnnyTeutonic commented Jun 4, 2017 • edited Loading

coveralls commented Jun 4, 2017

jni commented Jun 4, 2017

coveralls commented Jun 4, 2017

JohnnyTeutonic commented Jun 4, 2017 • edited Loading

jni commented Jun 4, 2017

jni commented Jun 4, 2017

JohnnyTeutonic commented Jun 4, 2017

jni commented Jun 4, 2017

coveralls commented Jun 4, 2017

JohnnyTeutonic commented Jun 4, 2017 •

edited

Loading

JohnnyTeutonic commented Jun 4, 2017 •

edited

Loading