[MRG] Fix dipole plot scale and smooth factors #730

kmilo9999 · 2024-03-15T18:47:11Z

Adding function to query if the data is a simulation or loaded from a file
Reorganizing the compare visualization tab
Adding scale and smooth factors float widgets
Fixing logic to apply the new parameters to the plots

solves #717

gtdang · 2024-03-15T21:13:36Z

hnn_core/gui/_viz_manager.py


-            return result
+                return result
+            except Exception as e:


Usually you specify what kind of exception you are trying to catch and handle here. Is there a specific exception that you are encountering?

Some GUI exceptions arent catch, like when out of bound or when it tries to use variables that haven't been set up yet.
This was my way to recognize the exception's message. Should I ditch these lines or swap them for something a bit more polished?

I see. If this is a good way for us to debug during development let's just keep this in for now.

is ipywidgets suppressing the errors? I seem to recall that all the errors were outputted to the console in the GUI. Is that not the case?

Maybe not all executions are logged to the console. Only script that is executed within a context manager of the logging window are logged. One of the first bugs I looked into was a plotting element that wasn't set within a context manager...

kmilo9999 · 2024-03-19T16:20:24Z

@ntolley @dylansdaniels
We have noticed that during the plotting of simulation data, the scaling and smooth factors (3000 and 30 respectively) are
automatically applied to the data. Those two default values are also applied to the loaded data (at plotting ) which modifies the value of the data. Also, when comparing the two datasets, both are adjusted by those factors which leads to the inaccuracies in the plot. To address this, we adjusted the initial values of the loaded data to have smooth factor of 0 and scaling factor 1 to preserve the original data values. This is the comparison plot between simulate and loaded data

What is the real purpose of the default smooth and scaling to 30 and 3000 respectively? What do you think about setting the loaded data default values to smooth = 0 and scaling = 1 ?

codecov-commenter · 2024-03-21T20:48:14Z

Codecov Report

Attention: Patch coverage is 87.50000% with 2 lines in your changes are missing coverage. Please review.

Project coverage is 92.73%. Comparing base (54ddd40) to head (118d2af).
Report is 1 commits behind head on master.

❗ Current head 118d2af differs from pull request most recent head e83b008. Consider uploading reports for the commit e83b008 to get more accurate results

Files	Patch %	Lines
hnn_core/gui/_viz_manager.py	86.66%	2 Missing ⚠️

❗ Your organization needs to install the Codecov GitHub app to enable full functionality.

Additional details and impacted files

@@            Coverage Diff             @@
##           master     #730      +/-   ##
==========================================
+ Coverage   92.66%   92.73%   +0.06%     
==========================================
  Files          27       27              
  Lines        4911     4917       +6     
==========================================
+ Hits         4551     4560       +9     
+ Misses        360      357       -3

☔ View full report in Codecov by Sentry.
📢 Have feedback on the report? Share it here.

…ields for data comparison

…ization

gtdang

Thanks Camilo! Left some comments.

gtdang · 2024-03-22T15:26:38Z

hnn_core/gui/_viz_manager.py

+    target_is_sim = True if single_simulation['net'] is not None else False
+    scaling = dipole_scaling.value if target_is_sim else data_scaling.value
+    smooth = dipole_smooth.value if target_is_sim else data_smooth.value
+    simulation_plot_config = {
        "max_spectral_frequency": max_spectral_frequency.value,
-        "dipole_scaling": dipole_scaling.value,
-        "dipole_smooth": dipole_smooth.value,
+        "dipole_scaling": scaling,
+        "dipole_smooth": smooth,
        "spectrogram_cm": spectrogram_colormap_selection.value
    }

    dpls_processed = _update_ax(fig, ax, single_simulation, sim_name,
-                                plot_type, plot_config)
+                                plot_type, simulation_plot_config)

    # If target_simulations is not None and we are plotting a dipole,
    # we need to plot the target dipole as well.
-    if target_simulations.value in data['simulations'].keys(
+    if data_widget.value in data['simulations'].keys(
    ) and plot_type == 'current dipole':

-        target_sim_name = target_simulations.value
+        target_sim_name = data_widget.value
        target_sim = data['simulations'][target_sim_name]
+        target_is_sim = True if target_sim['net'] is not None else False
+        scaling = dipole_scaling.value if target_is_sim else data_scaling.value
+        smooth = dipole_smooth.value if target_is_sim else data_smooth.value
+        data_plot_config = {
+            "max_spectral_frequency": max_spectral_frequency.value,
+            "dipole_scaling": scaling,
+            "dipole_smooth": smooth,
+            "spectrogram_cm": spectrogram_colormap_selection.value
+        }


Seems like repeat logic here. Perhaps create a function for compiling the data_plot_config dict based on if the data is a simulation or not?

I created a new function with the repeated code. Le me know your thoughts. I think it could improve

hnn_core/gui/_viz_manager.py

gtdang · 2024-03-22T15:38:08Z

hnn_core/gui/_viz_manager.py

+    tagert_names = simulation_names[0:-1]
+    if len(simulation_names) > 1:
+        tagert_names = simulation_names[1:]


Not sure I'm following here. It gets rid of the first simulation there is more than one set of data? Why?

There is a bug at creating a new plot using the "Make figure" button. The "data to compare" dropdown doesnt show all the available datasets to compare. They only show up when the value of the "Simulation Data" dropdown changes (a callback function is invoked and updates both lists). These lines initialize the list of available options without having to change the dropdowns values.

hnn_core/gui/_viz_manager.py

gtdang

Looking good Camilo. Just some minor edits.

hnn_core/gui/_viz_manager.py

hnn_core/gui/gui.py

gtdang

Looks great! Should be ready to merge once all checks pass!

jasmainak · 2024-03-26T14:04:52Z

hnn_core/gui/_viz_manager.py

-            # Unlink the widgets using the provided link object
-            link_attribute: link = getattr(self, attribute)
-            link_attribute.unlink()
+            try:


why do you need try-except? it's not a good idiom in Python ... replacing with if-else is preferable when possible

From Zen of Python:

Errors should never pass silently.
Unless explicitly silenced.

see also: https://peps.python.org/pep-0008/#programming-recommendations

There is internal code in the unlink call doesnt check the state of some attributes, causing out of bounds exceptions that are unhandled by hnn or the traitlets package. Additionally, the f handler call also throws exceptions unhandled by hnn. I can explicitly refer this exceptions in the try-except block or use if-else , but what if more exceptions come out? I am open to suggestions. Also, I can jut remove this block and create an issue for a future PR.

Camilo and I discussed this. It is useful as we are actively developing new features for the GUI over the next few months. Often times we're not debugging through a live GUI. We're going to keep it commented in pushes but will remove it entirely once the GUI is in production.

If you need it for debugging, it's fine. I would rather have it uncommented than not. But with an explanation why it's needed in the comments. I also realized that we're missing explanation for the unlink/relink logic in the code itself. I had to dig through github issues to understand why that exists. We should have a two-line comment with a pointer to an open github issue (on ipywidgets) suggesting when this logic can be removed. I worry we're adding patch on top of patch.

The try-except makes me nervous. It will suppress genuine errors. Please remove it and address what's necessary in a separate PR ... one never finds the time to go back and fix things.

I'm confused. This is actually going to print all errors and exceptions and not suppress any errors.

try: function() except Exception as e: print(f"{e}")

You could add a raise e if you want the program to stop upon finding an error. Though that's not ideal in the GUI as you will need to restart the GUI.

In production code you typically don't want to use except Exception, but something more specific like except ValueError, but that's only once you know the common exceptions that occur (and need to handle them in a specific way). The general Exception is good during development to find out what those common errors are.

Anyway, this logic is pretty minor and quick to add back while debugging so we should just remove it and not discuss it any further. But I did create issue #737 this week to remind CCV to review our work and remove any dev-related helper clauses/function before release that may slip through PRs.

Thanks :) The print statement isn't the same level as an error but it's certainly better than nothing. I would have preferred to have the errors thrown using the ipywidgets logging infrastructure as everything would then be visible in the same consolidated output panel. We can discuss more once I'm back from vacation. In the meantime, if you can inject the try-except as you need it, that would be great!

gtdang · 2024-03-29T18:15:09Z

Thanks Camilo!

* reorganizing visualization tab widgets. adding two additional float fields for data comparison * fixing gui test and funciton args name * fix linting errors * fix lint errors in gui.py * fixing gui tests * adding scaling and smooth input parameters for data comparison visualization * applied code review suggestions. Added functio to avoid repeated code * adding docustring to new funtion. removing unused is_loaded_data * fixing flake8 erros * removed logic to determine if data is a sim or loaded data * MAINT: commenting out try-except block * MAINT: uncommenting atry-except dev code in viz_manager * MAINT: remove try-except block

parent 289b85b author samadpls <[email protected]> 1710875966 +0500 committer samadpls <[email protected]> 1712146472 +0500 # This is a combination of 13 commits. parent 289b85b author samadpls <[email protected]> 1710875966 +0500 committer samadpls <[email protected]> 1712146405 +0500 # This is a combination of 4 commits.tree de20b388c64cca5a81e713c9270d6c1a3af12f93 parent 289b85b author samadpls <[email protected]> 1710875966 +0500 committer samadpls <[email protected]> 1712146383 +0500 # This is a combination of 3 commits. # This is the 1st commit message: Added `kwargs` options to `plot_spikes_hist` Signed-off-by: samadpls <[email protected]> Updated `whats_new.rst` with changes to plot_spikes_hist options using kwargs. Signed-off-by: samadpls <[email protected]> refactored Co-authored-by: Mainak Jas <[email protected]> refactored `kwargs_hist` Signed-off-by: samadpls <[email protected]> Updated `test_cell_response` with `kwargs_hist` Signed-off-by: samadpls <[email protected]> Update doc/whats_new.rst Co-authored-by: Mainak Jas <[email protected]> removed linewidth param Signed-off-by: GitHub <[email protected]> Added assertion Signed-off-by: samadpls <[email protected]> Added `kwargs` options to `plot_spikes_hist` Signed-off-by: samadpls <[email protected]> # This is the commit message #2: refactored Co-authored-by: Mainak Jas <[email protected]> # This is the commit message #3: refactored `kwargs_hist` Signed-off-by: samadpls <[email protected]> # This is the commit message jonescompneurolab#4: Update doc/whats_new.rst Co-authored-by: Mainak Jas <[email protected]> # This is the commit message jonescompneurolab#7: [MRG] Fix dipole plot scale and smooth factors (jonescompneurolab#730) * reorganizing visualization tab widgets. adding two additional float fields for data comparison * fixing gui test and funciton args name * fix linting errors * fix lint errors in gui.py * fixing gui tests * adding scaling and smooth input parameters for data comparison visualization * applied code review suggestions. Added functio to avoid repeated code * adding docustring to new funtion. removing unused is_loaded_data * fixing flake8 erros * removed logic to determine if data is a sim or loaded data * MAINT: commenting out try-except block * MAINT: uncommenting atry-except dev code in viz_manager * MAINT: remove try-except block # This is the commit message jonescompneurolab#8: [MRG] Added Virtual Environment Directories to `.gitignore` (jonescompneurolab#740) * Added virtual env * Updated * Refactored # This is the commit message jonescompneurolab#9: fix: correct pick_connection error where connections not relevant to the search criteria were returned when a search for non-existent connections was preformed # This is the commit message #10: style: clarify comments # This is the commit message jonescompneurolab#11: tests: added test for searching for a cell connection that is not configured in the network. # This is the commit message jonescompneurolab#12: style: flake8 edits # This is the commit message #13: refactor: convert to lists # This is the commit message jonescompneurolab#15: chore: Corrected whats_new.rst tense. # This is the commit message jonescompneurolab#17: refactor: simplified logic # This is the commit message #18: reorganizing visualization tab widgets. adding two additional float fields for data comparison fix linting errors fix lint errors in gui.py adding scaling and smooth input parameters for data comparison visualization applied code review suggestions. Added functio to avoid repeated code adding docustring to new funtion. removing unused is_loaded_data MAINT: commenting out try-except block MAINT: uncommenting atry-except dev code in viz_manager

kmilo9999 requested a review from gtdang March 15, 2024 18:47

gtdang reviewed Mar 15, 2024

View reviewed changes

gtdang requested review from dylansdaniels and ntolley March 18, 2024 19:40

gtdang linked an issue Mar 18, 2024 that may be closed by this pull request

BUG (GUI): Dipole plot scaling #717

Closed

gtdang changed the title ~~Fix dipole plot scale and smooth factors~~ [WIP] Fix dipole plot scale and smooth factors Mar 18, 2024

kmilo9999 marked this pull request as ready for review March 22, 2024 13:33

kmilo9999 added 6 commits March 22, 2024 09:45

reorganizing visualization tab widgets. adding two additional float f…

5f235be

…ields for data comparison

fixing gui test and funciton args name

95ed9c1

fix linting errors

3ac1386

fix lint errors in gui.py

ccb255a

fixing gui tests

d480947

adding scaling and smooth input parameters for data comparison visual…

affbfda

…ization

kmilo9999 force-pushed the fix-dipole-fig-scale branch from d03a7ca to affbfda Compare March 22, 2024 14:39

kmilo9999 requested a review from gtdang March 22, 2024 15:41

gtdang requested changes Mar 22, 2024

View reviewed changes

applied code review suggestions. Added functio to avoid repeated code

e47a1e8

gtdang self-requested a review March 22, 2024 18:39

gtdang requested changes Mar 22, 2024

View reviewed changes

hnn_core/gui/_viz_manager.py Outdated Show resolved Hide resolved

hnn_core/gui/_viz_manager.py Show resolved Hide resolved

hnn_core/gui/gui.py Outdated Show resolved Hide resolved

adding docustring to new funtion. removing unused is_loaded_data

e3ae1c4

kmilo9999 requested a review from gtdang March 22, 2024 22:20

kmilo9999 added 2 commits March 22, 2024 20:43

fixing flake8 erros

c4f2319

removed logic to determine if data is a sim or loaded data

3be7a2d

gtdang reviewed Mar 26, 2024

View reviewed changes

gtdang approved these changes Mar 26, 2024

View reviewed changes

jasmainak reviewed Mar 26, 2024

View reviewed changes

kmilo9999 added 2 commits March 26, 2024 12:17

MAINT: commenting out try-except block

7de6d8f

MAINT: uncommenting atry-except dev code in viz_manager

2dcf972

kmilo9999 changed the title ~~[WIP] Fix dipole plot scale and smooth factors~~ [MRG] Fix dipole plot scale and smooth factors Mar 27, 2024

kmilo9999 mentioned this pull request Mar 29, 2024

GUI: Creating figures with no data throws multiple exceptions #742

Closed

MAINT: remove try-except block

e83b008

gtdang merged commit 19aff25 into jonescompneurolab:master Mar 29, 2024
9 checks passed

gtdang mentioned this pull request Mar 29, 2024

[MRG] Fix RMSE Annotation #741

Merged

gtdang deleted the fix-dipole-fig-scale branch August 29, 2024 17:27

Provide feedback

Saved searches

Use saved searches to filter your results more quickly

[MRG] Fix dipole plot scale and smooth factors #730

[MRG] Fix dipole plot scale and smooth factors #730

kmilo9999 commented Mar 15, 2024 •

edited

Loading

gtdang Mar 15, 2024

kmilo9999 Mar 22, 2024

gtdang Mar 22, 2024

jasmainak Mar 26, 2024

gtdang Mar 26, 2024

kmilo9999 commented Mar 19, 2024

codecov-commenter commented Mar 21, 2024 •

edited

Loading

gtdang left a comment

gtdang Mar 22, 2024

kmilo9999 Mar 22, 2024

gtdang Mar 22, 2024

kmilo9999 Mar 22, 2024

gtdang left a comment

gtdang left a comment

jasmainak Mar 26, 2024

jasmainak Mar 26, 2024

kmilo9999 Mar 26, 2024

gtdang Mar 26, 2024

jasmainak Mar 27, 2024 •

edited

Loading

kmilo9999 Mar 29, 2024

jasmainak Mar 29, 2024

gtdang Mar 29, 2024 •

edited

Loading

gtdang Mar 29, 2024

jasmainak Mar 29, 2024

gtdang commented Mar 29, 2024

[MRG] Fix dipole plot scale and smooth factors #730

[MRG] Fix dipole plot scale and smooth factors #730

Conversation

kmilo9999 commented Mar 15, 2024 • edited Loading

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

kmilo9999 commented Mar 19, 2024

codecov-commenter commented Mar 21, 2024 • edited Loading

Codecov Report

gtdang 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

gtdang left a comment

Choose a reason for hiding this comment

gtdang 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

jasmainak Mar 27, 2024 • edited Loading

Choose a reason for hiding this comment

Choose a reason for hiding this comment

Choose a reason for hiding this comment

gtdang Mar 29, 2024 • edited Loading

Choose a reason for hiding this comment

Choose a reason for hiding this comment

Choose a reason for hiding this comment

gtdang commented Mar 29, 2024

kmilo9999 commented Mar 15, 2024 •

edited

Loading

codecov-commenter commented Mar 21, 2024 •

edited

Loading

jasmainak Mar 27, 2024 •

edited

Loading

gtdang Mar 29, 2024 •

edited

Loading