Experiment Class Refactor (update to #183), converting specific experiments to subclasses #184

Parvfect · 2022-06-15T12:09:18Z

As per the comments left in PR #183, Sub classes make more sense for increasing modular support and as a general framework. I've implemented the base Experiment Class and the Experiment Subclasses for Visual P300, N170 and SSVEP (although I can rewrite some of the code in SSVEP and make it better). Next step is running it and making sure it works the same, I'm having some trouble in downloading the psychopy package due to my Python version and I should get that done very soon. Wanted to put this through so we could talk about it in the dev meeting.

Changes as per the project proposal for Google Summer of Code for eeg-notebooks - https://summerofcode.withgoogle.com/programs/2022/projects/2baQ46dm.

Implemented as a part of the GSOC period for EEG - Notebooks under the mentorship of John Griffiths.

…into experimentSubclass2

Parvfect · 2022-06-27T22:30:10Z

Right so the Visual P300 and the Visual N170 works although its a bit different,

'''
from eegnb.experiments.visual_n170 import n170
t = n170.VisualN170()
t.present()
''''

That should run it.

A few things worth pointing out,

I have changed how the windows work, previously two mywin instances were used to show instructions and present whereas I have reused the first one - don't know if this is ideal and need some advice.
I have a shit ton of self parameters to share variables across functions and I don't know if that's fun for someone new who tries to create an experiment from this structure -> might need to be discussed
There are some bits of ugly code here and there that I am going to clean in the next few days
Some inputs in some of the files seem redundant if it's importing the Experiment Class

JohnGriffiths

Overall: like this a lot. Seems to run successfully from tests I have done so far. New structure is good.

Several specific suggestions and requests in the code.

General request:

A lot more in-code comments + annotation + documentation.

JohnGriffiths · 2022-06-30T22:21:07Z

eegnb/experiments/Experiment.py

+    def present(self):
+        """ Do the present operation for a bunch of experiments """
+
+        self.setup()


As discussed, let's add an argument here so that instructions can optionally be skipped. Default behaviour should be to show however.

JohnGriffiths · 2022-06-30T22:24:02Z

eegnb/experiments/Experiment.py

+            # Offset
+            core.wait(self.soa)
+            self.mywin.flip()
+            if len(event.getKeys()) > 0 or (time() - start) > self.record_duration:


Also as discussed, this part (of the original code) has always looked a bit ugly to me. The alternative is to pre-calculate all of the itis, soas, jitters in advance, and then it will be known before the loop starts what the duration will be, and so the decision of whether or it is the duration param or the completion of the trial set will be clear and not determined on the fly at the end.

However, I suggest we have a stab at this small adjustment at a later point, it isn't the focus or priority of the current PR.

JohnGriffiths · 2022-06-30T22:30:19Z

eegnb/experiments/visual_n170/n170.py

@@ -17,107 +17,39 @@
 from eegnb import generate_save_fn
 from eegnb.devices.eeg import EEG
 from eegnb.stimuli import FACE_HOUSE
+from eegnb.experiments.Experiment import Experiment


having both the file name and the module name the exact same text (Experiment) may be common but is not clean.

I think the experiment base class should be called something like `BaseExperiment'

Also I think we should import the final experiment classes into the init so that they can be imported more cleanly at the user level.

So the final use pattern would be something like

from eegnb.experiments import VisualN170 expt = VisualN170() expt.present()

A problem with this is that the file VisualN170 is super buried in the folders, I don't mind bringing all the experiments out of their specific folder so this is possible but need to make sure it does not cause any other problems.

JohnGriffiths · 2022-06-30T22:31:34Z

eegnb/experiments/Experiment.py

+from eegnb import generate_save_fn
+from eegnb.devices.eeg import EEG
+
+class Experiment:


see comment below. I think

class ExperimentBase

or

class BaseExperiment

If the other experiments are named P300Experiment for example, then it should prob be BaseExperiment

I agree with you guys, gonna change it to BaseExperiment
I faced some namespace errors when trying to get rid of the redundant imports so this would need to be done anyway.

JohnGriffiths · 2022-06-30T22:33:44Z

eegnb/experiments/Experiment.py

+                    f"No path for a save file was passed to the experiment. Saving data to {self.save_fn}"
+                )
+
+    def present(self):


I actually don't like the name present for this method. Never did.

I think this 'run experiment' method should be

def run(self):

this would also help avoid any confusion with the present_stimulus method

ErikBjare

Looking good! Just had a few naming nits :)

ErikBjare · 2022-07-04T13:59:53Z

eegnb/experiments/Experiment.py

-        mywin = visual.Window([1600, 900], monitor="testMonitor", units="deg", fullscr=True)
-        mywin.mouseVisible = False
+        #mywin = visual.Window([1600, 900], monitor="testMonitor", units="deg", fullscr=True)
+        self.mywin.mouseVisible = False


Name it window instead of mywin

ErikBjare · 2022-07-04T14:01:52Z

eegnb/experiments/visual_p300/p300.py


-    def present_stimulus(self):
+    def present_stimulus(self, ii):


I think idx is a better name than ii (which I've seen used for vectors), type annotation would also help :)

I agree, done

Parvfect · 2022-07-06T11:11:27Z

Have adressed the comments given above and made the naming changes.
Two issues still remaining

For importing experiments directly the files will have to be removed from their folders that is from (VisualN170/n170.py) to just the file. Don't think this should cause any complications but need to make sure this is what is needed .
I can't seem to get rid of the redundant imports in the specific experiments by putting in the BaseExperiment class, if anyone can help with that it would be lovely.

Besides that I think I should be good to go and start changing the other Experiments for this style, just need a nod from John to start it.

JohnGriffiths · 2022-07-21T16:18:05Z

Getting error with this.

Running

eegnb runexp -ip muse2_bfn

Gives

Traceback (most recent call last):
  File "C:\Users\eeg_lab\.conda\envs\eeg-notebooks\Scripts\eegnb-script.py", line 33, in <module>
    sys.exit(load_entry_point('eeg-notebooks', 'console_scripts', 'eegnb')())
  File "C:\Users\eeg_lab\.conda\envs\eeg-notebooks\lib\site-packages\click\core.py", line 1128, in __call__
    return self.main(*args, **kwargs)
  File "C:\Users\eeg_lab\.conda\envs\eeg-notebooks\lib\site-packages\click\core.py", line 1053, in main
    rv = self.invoke(ctx)
  File "C:\Users\eeg_lab\.conda\envs\eeg-notebooks\lib\site-packages\click\core.py", line 1659, in invoke
    return _process_result(sub_ctx.command.invoke(sub_ctx))
  File "C:\Users\eeg_lab\.conda\envs\eeg-notebooks\lib\site-packages\click\core.py", line 1395, in invoke
    return ctx.invoke(self.callback, **ctx.params)
  File "C:\Users\eeg_lab\.conda\envs\eeg-notebooks\lib\site-packages\click\core.py", line 754, in invoke
    return __callback(*args, **kwargs)
  File "c:\users\eeg_lab\code\github\eeg-notebooks_dev\eeg-notebooks_reviewpr184\eegnb\cli\__main__.py", line 80, in runexp
    run_experiment(experiment, eeg, recdur, outfname)
  File "c:\users\eeg_lab\code\github\eeg-notebooks_dev\eeg-notebooks_reviewpr184\eegnb\cli\utils.py", line 45, in run_experiment
    module.present(duration=record_duration, eeg=eeg_device, save_fn=save_fn)  # type: ignore
AttributeError: module 'eegnb.experiments.visual_n170.n170' has no attribute 'present'

Parvfect · 2022-07-29T00:12:44Z

Getting error with this.

Running

eegnb runexp -ip muse2_bfn

Gives

Traceback (most recent call last):
  File "C:\Users\eeg_lab\.conda\envs\eeg-notebooks\Scripts\eegnb-script.py", line 33, in <module>
    sys.exit(load_entry_point('eeg-notebooks', 'console_scripts', 'eegnb')())
  File "C:\Users\eeg_lab\.conda\envs\eeg-notebooks\lib\site-packages\click\core.py", line 1128, in __call__
    return self.main(*args, **kwargs)
  File "C:\Users\eeg_lab\.conda\envs\eeg-notebooks\lib\site-packages\click\core.py", line 1053, in main
    rv = self.invoke(ctx)
  File "C:\Users\eeg_lab\.conda\envs\eeg-notebooks\lib\site-packages\click\core.py", line 1659, in invoke
    return _process_result(sub_ctx.command.invoke(sub_ctx))
  File "C:\Users\eeg_lab\.conda\envs\eeg-notebooks\lib\site-packages\click\core.py", line 1395, in invoke
    return ctx.invoke(self.callback, **ctx.params)
  File "C:\Users\eeg_lab\.conda\envs\eeg-notebooks\lib\site-packages\click\core.py", line 754, in invoke
    return __callback(*args, **kwargs)
  File "c:\users\eeg_lab\code\github\eeg-notebooks_dev\eeg-notebooks_reviewpr184\eegnb\cli\__main__.py", line 80, in runexp
    run_experiment(experiment, eeg, recdur, outfname)
  File "c:\users\eeg_lab\code\github\eeg-notebooks_dev\eeg-notebooks_reviewpr184\eegnb\cli\utils.py", line 45, in run_experiment
    module.present(duration=record_duration, eeg=eeg_device, save_fn=save_fn)  # type: ignore
AttributeError: module 'eegnb.experiments.visual_n170.n170' has no attribute 'present'

So this is because we changed our present function to run. Additionally, the runexp script will need to be modified because the usage is not eegnb.experiments.visual_n170.n170 anymore and we have to initalize a class object first of the form,

'from eegnb.devices.eeg import EEG
from eegnb.experiments import VisualN170, VisualP300, VisualSSVEP, AuditoryOddball

if name == "main":
board_name = 'muse2'

record_duration = 120
eeg_device = EEG(device=board_name)
t = AuditoryOddball(duration=record_duration, eeg=eeg_device)
t.run()

'

I'm not too sure where the runexp file is, I'll try finiding it right now so I can make the necessary changes.

Parvfect · 2022-07-29T00:15:47Z

Okay so I tested each experiment changed (VisualN170, VisualP300, VisualSSVEP and AuditoryOddball) using the Muse2 and a script importing them directly rather than through runexp. They all seem to be functional, no errors and running smooth. There seems to be a slight delay from the spacebar of the instruction screen to the displaying of images, which I assume is setting up the eeg-stream, but it's definitely longer than normal and I'll need another pair of eyes on it to see if it's a problem.

As of now, I need to fix the cli usage of it as John has shown above and more robust tests for making sure everything is working as we need. I also have not been able to test this through the Brainflow, have been using Bluemuse.

Parvfect · 2022-08-03T15:45:55Z

Getting error with this.
Running

eegnb runexp -ip muse2_bfn

Gives

Traceback (most recent call last):
  File "C:\Users\eeg_lab\.conda\envs\eeg-notebooks\Scripts\eegnb-script.py", line 33, in <module>
    sys.exit(load_entry_point('eeg-notebooks', 'console_scripts', 'eegnb')())
  File "C:\Users\eeg_lab\.conda\envs\eeg-notebooks\lib\site-packages\click\core.py", line 1128, in __call__
    return self.main(*args, **kwargs)
  File "C:\Users\eeg_lab\.conda\envs\eeg-notebooks\lib\site-packages\click\core.py", line 1053, in main
    rv = self.invoke(ctx)
  File "C:\Users\eeg_lab\.conda\envs\eeg-notebooks\lib\site-packages\click\core.py", line 1659, in invoke
    return _process_result(sub_ctx.command.invoke(sub_ctx))
  File "C:\Users\eeg_lab\.conda\envs\eeg-notebooks\lib\site-packages\click\core.py", line 1395, in invoke
    return ctx.invoke(self.callback, **ctx.params)
  File "C:\Users\eeg_lab\.conda\envs\eeg-notebooks\lib\site-packages\click\core.py", line 754, in invoke
    return __callback(*args, **kwargs)
  File "c:\users\eeg_lab\code\github\eeg-notebooks_dev\eeg-notebooks_reviewpr184\eegnb\cli\__main__.py", line 80, in runexp
    run_experiment(experiment, eeg, recdur, outfname)
  File "c:\users\eeg_lab\code\github\eeg-notebooks_dev\eeg-notebooks_reviewpr184\eegnb\cli\utils.py", line 45, in run_experiment
    module.present(duration=record_duration, eeg=eeg_device, save_fn=save_fn)  # type: ignore
AttributeError: module 'eegnb.experiments.visual_n170.n170' has no attribute 'present'

So this is because we changed our present function to run. Additionally, the runexp script will need to be modified because the usage is not eegnb.experiments.visual_n170.n170 anymore and we have to initalize a class object first of the form,

'from eegnb.devices.eeg import EEG from eegnb.experiments import VisualN170, VisualP300, VisualSSVEP, AuditoryOddball

if name == "main": board_name = 'muse2'

record_duration = 120
eeg_device = EEG(device=board_name)
t = AuditoryOddball(duration=record_duration, eeg=eeg_device)
t.run()

'

I'm not too sure where the runexp file is, I'll try finiding it right now so I can make the necessary changes.

Fixed this in new commit. Please note that an if condition had to be added for the new calling style that can be made cleaner if all the experiments are adapted.

Code looks like this (eegnb/cli/utils.py) for initialization

from eegnb.experiments import VisualN170
from eegnb.experiments import VisualP300
from eegnb.experiments import VisualSSVEP
from eegnb.experiments import AuditoryOddball
from eegnb.experiments.visual_cueing import cueing
from eegnb.experiments.visual_codeprose import codeprose
from eegnb.experiments.auditory_oddball import diaconescu
from eegnb.experiments.auditory_ssaep import ssaep, ssaep_onefreq


# New Experiment Class structure has a different initilization, to be noted
experiments = {
    "visual-N170": VisualN170(),
    "visual-P300": VisualP300(),
    "visual-SSVEP": VisualSSVEP(),
    "visual-cue": cueing,
    "visual-codeprose": codeprose,
    "auditory-SSAEP orig": ssaep,
    "auditory-SSAEP onefreq": ssaep_onefreq,
    "auditory-oddball orig": AuditoryOddball(),
    "auditory-oddball diaconescu": diaconescu,
}

and for the running,

def run_experiment(
    experiment: str, eeg_device: EEG, record_duration: float = None, save_fn=None
):
    if experiment in experiments:
        module = experiments[experiment]

        # Condition added for different run types of old and new experiment class structure
        if experiment == "visual-N170" or experiment == "visual-P300" or experiment == "visual-SSVEP" or experiment == "auditory-oddball orig":
            module.duration = record_duration
            module.eeg = eeg_device
            module.save_fn = save_fn
            module.run()
        else:
            module.present(duration=record_duration, eeg=eeg_device, save_fn=save_fn)  # type: ignore
    else:
        print("\nError: Unknown experiment '{}'".format(experiment))
        print("\nExperiment can be one of:")
        print("\n".join([" - " + exp for exp in experiments]))

Parvfect · 2022-08-04T19:42:15Z

For Reviewer

Test each experiment (Visual N170, Visual P300, Auditory Oddball, Visual SSVEP) using the CLI
using the EEG (already working without)

eegnb runexp -ip

There seems to be a slight delay between the instruction screen and running the experiments that is bothering me, really need someone to confirm that it's alright. Besides that everything else seems alright.

JohnGriffiths · 2022-08-04T20:46:31Z

This isn't working for me. I get 1 image in the sequence shown, then it closes down immediately. Tried for Muse 2016 and Muse S on Windows 10 ( I think). Trying to get to the bottom of it.

Parvfect · 2022-08-05T10:41:52Z

This isn't working for me. I get 1 image in the sequence shown, then it closes down immediately. Tried for Muse 2016 and Muse S on Windows 10 ( I think). Trying to get to the bottom of it.

Send me any error messages or outputs if you can, could look into it and find out more.

JohnGriffiths · 2022-08-09T18:00:06Z

Ok, apols for false alarm. This appears to be working ok for me now. Tested for museS and muse2016 on two windows 10 comps, with brainflow native device option.

I agree that the wait period between pressing space and the experiment starting seems a bit long. I wonder if that can be sped up?

If that is difficult then the message could be adjusted to something like

'please press space and experiment will begin in a few seconds'

Haven't yet checked that the ERPs recorded with this look ok. No real reason why they shouldn't but is obvs super important that that is confirmed. Will get to that asap.

Will do a final review of code changes shortly, and if all is good then I am good to merge this PR.

JohnGriffiths · 2022-08-10T16:34:50Z

Merging to dev.

ErikBjare

Nice to see this merged! Just had a few comments.

ErikBjare · 2022-08-11T09:45:55Z

eegnb/cli/utils.py

+        if experiment == "visual-N170" or experiment == "visual-P300" or experiment == "visual-SSVEP" or experiment == "auditory-oddball orig":
+            module.duration = record_duration
+            module.eeg = eeg_device
+            module.save_fn = save_fn
+            module.run()


Should be replaced with:

if isinstance(experiment, BaseExperiment.__class__): exp = experiment(eeg=eeg, duration=record_duration, save_fn=save_fn) exp.run()

ErikBjare · 2022-08-11T09:45:57Z

eegnb/cli/utils.py

 experiments = {
-    "visual-N170": n170,
-    "visual-P300": p300,
-    "visual-SSVEP": ssvep,
+    "visual-N170": VisualN170(),
+    "visual-P300": VisualP300(),
+    "visual-SSVEP": VisualSSVEP(),
    "visual-cue": cueing,
    "visual-codeprose": codeprose,
    "auditory-SSAEP orig": ssaep,
    "auditory-SSAEP onefreq": ssaep_onefreq,
-    "auditory-oddball orig": aob,
+    "auditory-oddball orig": AuditoryOddball(),
    "auditory-oddball diaconescu": diaconescu,


These should not be instantiated globally, just leave the class in the dict and let the caller handle initialization.

So that, for example:

"visual-N170": VisualN170(),

Just becomes

"visual-N170": VisualN170,

ErikBjare · 2022-08-11T09:48:42Z

eegnb/experiments/Experiment.py

+prefs.hardware['audioLib'] = 'PTB'
+prefs.hardware['audioLatencyMode'] = 3


Are these safe to set globally? @JohnGriffiths

Feels like we set them like this in more places, in which case they would conflict.

Edit: Indeed, but looks like they all set the same values? Should prob be consolidated. All hits: https://github.com/NeuroTechX/eeg-notebooks/search?q=prefs.hardware

…ific experiments to subclasses (NeuroTechX#184) * First commit * Second commit * Modifications * Lol * Lol * Incorporated N170 and p300, looking good for a PR * ssvep update * Implementing subclasses instead of loose functions * fix: fixed import (brainflow updated API) * Playing around still * Fixing import errors * Adding abstractmethod decorators * Still working on the import error * Guess what's finally working * Comments and naming ticks * More comments * Live coding demonstration * ssvep adapted * Adapting Auditory Oddball * changing save_fn to self.save_fun * This maybe the last big change * utils file changed, changes work through cli Co-authored-by: Erik Bjäreholt <[email protected]>

* example test commit (#182) * example test commit * example edit * ci: run test workflow on develop branch * ci: add develop branch to job triggers * ci: fix syntax issue in workflow * fix: fixed import (brainflow updated API) * build(deps): locked pylsl==1.10.5 (#187) * Experiment Class Refactor (update to #183), converting specific experiments to subclasses (#184) * First commit * Second commit * Modifications * Lol * Lol * Incorporated N170 and p300, looking good for a PR * ssvep update * Implementing subclasses instead of loose functions * fix: fixed import (brainflow updated API) * Playing around still * Fixing import errors * Adding abstractmethod decorators * Still working on the import error * Guess what's finally working * Comments and naming ticks * More comments * Live coding demonstration * ssvep adapted * Adapting Auditory Oddball * changing save_fn to self.save_fun * This maybe the last big change * utils file changed, changes work through cli Co-authored-by: Erik Bjäreholt <[email protected]> * Submodule added for gsoc * Adding pipelines for cli analysis (#202) * started pipelines function * almost working simple function equivalents of nb scripts * fix: fixed import (brainflow updated API) * sqc fixes for unicorn (#176) * Ignore pushes * Trying to create a cli * Stepping through the problem * First commit * Fixing pause in signal quality check * Fixing Signal quality check problem * fix the technical debt * Save path done for automated saving pdf * I feel amazing * Almost through * Update eegnb/cli/__main__.py Co-authored-by: Erik Bjäreholt <[email protected]> * Trying to create cli but it's being really painful * Extra word cli error * Changed example handling * Pain * Adding whole datapath * Finally fixed cli * hmm * Looking good * added hyperlink * Having some issues with detecting css and image deltetion * Just the css now * Fixed the css linking problem though it's a weird soln * Automated running, still fnames problem * Hahahah embedded images in html * Improving code * Okay now * Look at that * Almost there just the two figures now * Now * Added attrdict to do with cli error Co-authored-by: John Griffiths <[email protected]> Co-authored-by: Erik Bjäreholt <[email protected]> Co-authored-by: John Griffiths <[email protected]> * added more options for site args; improved function names; removed some redundant lines (#209) * fix subject num parsing bug * analysis report function improvements for openbci cyton and gtec unicorn devices * run exp fix * Update requirements.txt * fixes to get docs building by github action (#210) * fixes to get docs building by github action * reverted some changes * Update 01r__ssvep_viz.py Co-authored-by: John Griffiths <[email protected]> * Update README.rst small commit to test doc build workflow on this branch * removing gsoc submodule Co-authored-by: Erik Bjäreholt <[email protected]> Co-authored-by: Parv Agarwal <[email protected]> Co-authored-by: Parvfect <[email protected]> Co-authored-by: Ben Pettit <[email protected]>

* major update: merging develop to master (#217) * example test commit (#182) * example test commit * example edit * ci: run test workflow on develop branch * ci: add develop branch to job triggers * ci: fix syntax issue in workflow * fix: fixed import (brainflow updated API) * build(deps): locked pylsl==1.10.5 (#187) * Experiment Class Refactor (update to #183), converting specific experiments to subclasses (#184) * First commit * Second commit * Modifications * Lol * Lol * Incorporated N170 and p300, looking good for a PR * ssvep update * Implementing subclasses instead of loose functions * fix: fixed import (brainflow updated API) * Playing around still * Fixing import errors * Adding abstractmethod decorators * Still working on the import error * Guess what's finally working * Comments and naming ticks * More comments * Live coding demonstration * ssvep adapted * Adapting Auditory Oddball * changing save_fn to self.save_fun * This maybe the last big change * utils file changed, changes work through cli Co-authored-by: Erik Bjäreholt <[email protected]> * Submodule added for gsoc * Adding pipelines for cli analysis (#202) * started pipelines function * almost working simple function equivalents of nb scripts * fix: fixed import (brainflow updated API) * sqc fixes for unicorn (#176) * Ignore pushes * Trying to create a cli * Stepping through the problem * First commit * Fixing pause in signal quality check * Fixing Signal quality check problem * fix the technical debt * Save path done for automated saving pdf * I feel amazing * Almost through * Update eegnb/cli/__main__.py Co-authored-by: Erik Bjäreholt <[email protected]> * Trying to create cli but it's being really painful * Extra word cli error * Changed example handling * Pain * Adding whole datapath * Finally fixed cli * hmm * Looking good * added hyperlink * Having some issues with detecting css and image deltetion * Just the css now * Fixed the css linking problem though it's a weird soln * Automated running, still fnames problem * Hahahah embedded images in html * Improving code * Okay now * Look at that * Almost there just the two figures now * Now * Added attrdict to do with cli error Co-authored-by: John Griffiths <[email protected]> Co-authored-by: Erik Bjäreholt <[email protected]> Co-authored-by: John Griffiths <[email protected]> * added more options for site args; improved function names; removed some redundant lines (#209) * fix subject num parsing bug * analysis report function improvements for openbci cyton and gtec unicorn devices * run exp fix * Update requirements.txt * fixes to get docs building by github action (#210) * fixes to get docs building by github action * reverted some changes * Update 01r__ssvep_viz.py Co-authored-by: John Griffiths <[email protected]> * Update README.rst small commit to test doc build workflow on this branch * removing gsoc submodule Co-authored-by: Erik Bjäreholt <[email protected]> Co-authored-by: Parv Agarwal <[email protected]> Co-authored-by: Parvfect <[email protected]> Co-authored-by: Ben Pettit <[email protected]> * update dependencies - seaborn * docs/perf: reduced the imports: `cueing` example * bug: update deprecated `plot_psd()` method - feature of new `mne` version. - instead of doing plot_psd() from the `mne.io.Raw` object, must do this: - `raw.compute_psd().plot()` - i.e., has to pass through a `spectrum` object * updated deprec. `mne` function * perf: removed importage of unused packages from example - One of them, i.e., `collections.Iterable` is even deprecated. - Must use `collections.abc.Iterable` instead now. - Resulting in faster build/user run * bugfix: `plot_conditions` - due to `sns` deprecation * bugfix: resolved `psd_welch()` deprecation (`mne`) --------- Co-authored-by: John Griffiths <[email protected]> Co-authored-by: Erik Bjäreholt <[email protected]> Co-authored-by: Parv Agarwal <[email protected]> Co-authored-by: Parvfect <[email protected]> Co-authored-by: Ben Pettit <[email protected]> Co-authored-by: Taha Morshedzadeh <[email protected]>

…Refactor (#218) * major update: merging develop to master (#217) * example test commit (#182) * example test commit * example edit * ci: run test workflow on develop branch * ci: add develop branch to job triggers * ci: fix syntax issue in workflow * fix: fixed import (brainflow updated API) * build(deps): locked pylsl==1.10.5 (#187) * Experiment Class Refactor (update to #183), converting specific experiments to subclasses (#184) * First commit * Second commit * Modifications * Lol * Lol * Incorporated N170 and p300, looking good for a PR * ssvep update * Implementing subclasses instead of loose functions * fix: fixed import (brainflow updated API) * Playing around still * Fixing import errors * Adding abstractmethod decorators * Still working on the import error * Guess what's finally working * Comments and naming ticks * More comments * Live coding demonstration * ssvep adapted * Adapting Auditory Oddball * changing save_fn to self.save_fun * This maybe the last big change * utils file changed, changes work through cli Co-authored-by: Erik Bjäreholt <[email protected]> * Submodule added for gsoc * Adding pipelines for cli analysis (#202) * started pipelines function * almost working simple function equivalents of nb scripts * fix: fixed import (brainflow updated API) * sqc fixes for unicorn (#176) * Ignore pushes * Trying to create a cli * Stepping through the problem * First commit * Fixing pause in signal quality check * Fixing Signal quality check problem * fix the technical debt * Save path done for automated saving pdf * I feel amazing * Almost through * Update eegnb/cli/__main__.py Co-authored-by: Erik Bjäreholt <[email protected]> * Trying to create cli but it's being really painful * Extra word cli error * Changed example handling * Pain * Adding whole datapath * Finally fixed cli * hmm * Looking good * added hyperlink * Having some issues with detecting css and image deltetion * Just the css now * Fixed the css linking problem though it's a weird soln * Automated running, still fnames problem * Hahahah embedded images in html * Improving code * Okay now * Look at that * Almost there just the two figures now * Now * Added attrdict to do with cli error Co-authored-by: John Griffiths <[email protected]> Co-authored-by: Erik Bjäreholt <[email protected]> Co-authored-by: John Griffiths <[email protected]> * added more options for site args; improved function names; removed some redundant lines (#209) * fix subject num parsing bug * analysis report function improvements for openbci cyton and gtec unicorn devices * run exp fix * Update requirements.txt * fixes to get docs building by github action (#210) * fixes to get docs building by github action * reverted some changes * Update 01r__ssvep_viz.py Co-authored-by: John Griffiths <[email protected]> * Update README.rst small commit to test doc build workflow on this branch * removing gsoc submodule Co-authored-by: Erik Bjäreholt <[email protected]> Co-authored-by: Parv Agarwal <[email protected]> Co-authored-by: Parvfect <[email protected]> Co-authored-by: Ben Pettit <[email protected]> * Updated doc examples * Update 00x__n170_run_experiment.py fix: typo in func param --------- Co-authored-by: John Griffiths <[email protected]> Co-authored-by: Erik Bjäreholt <[email protected]> Co-authored-by: Ben Pettit <[email protected]> Co-authored-by: Ore O <[email protected]>

* example test commit (#182) * example test commit * example edit * ci: run test workflow on develop branch * ci: add develop branch to job triggers * ci: fix syntax issue in workflow * fix: fixed import (brainflow updated API) * build(deps): locked pylsl==1.10.5 (#187) * Experiment Class Refactor (update to #183), converting specific experiments to subclasses (#184) * First commit * Second commit * Modifications * Lol * Lol * Incorporated N170 and p300, looking good for a PR * ssvep update * Implementing subclasses instead of loose functions * fix: fixed import (brainflow updated API) * Playing around still * Fixing import errors * Adding abstractmethod decorators * Still working on the import error * Guess what's finally working * Comments and naming ticks * More comments * Live coding demonstration * ssvep adapted * Adapting Auditory Oddball * changing save_fn to self.save_fun * This maybe the last big change * utils file changed, changes work through cli Co-authored-by: Erik Bjäreholt <[email protected]> * Submodule added for gsoc * Adding pipelines for cli analysis (#202) * started pipelines function * almost working simple function equivalents of nb scripts * fix: fixed import (brainflow updated API) * sqc fixes for unicorn (#176) * Ignore pushes * Trying to create a cli * Stepping through the problem * First commit * Fixing pause in signal quality check * Fixing Signal quality check problem * fix the technical debt * Save path done for automated saving pdf * I feel amazing * Almost through * Update eegnb/cli/__main__.py Co-authored-by: Erik Bjäreholt <[email protected]> * Trying to create cli but it's being really painful * Extra word cli error * Changed example handling * Pain * Adding whole datapath * Finally fixed cli * hmm * Looking good * added hyperlink * Having some issues with detecting css and image deltetion * Just the css now * Fixed the css linking problem though it's a weird soln * Automated running, still fnames problem * Hahahah embedded images in html * Improving code * Okay now * Look at that * Almost there just the two figures now * Now * Added attrdict to do with cli error Co-authored-by: John Griffiths <[email protected]> Co-authored-by: Erik Bjäreholt <[email protected]> Co-authored-by: John Griffiths <[email protected]> * added more options for site args; improved function names; removed some redundant lines (#209) * fix subject num parsing bug * analysis report function improvements for openbci cyton and gtec unicorn devices * run exp fix * Update requirements.txt * fixes to get docs building by github action (#210) * fixes to get docs building by github action * reverted some changes * Update 01r__ssvep_viz.py Co-authored-by: John Griffiths <[email protected]> * Update README.rst small commit to test doc build workflow on this branch * removing gsoc submodule * ci: update python to 3.8, update wxPython, update setup-python action * ci: pin ubuntu versions to 22.04, updated wxPython urls * ci: bumped psychopy to 2023.1.0 * build(deps): set upper supported numpy version to 1.23.x * chore: applied no_implicit_optional * update dependencies - for build (#220) * major update: merging develop to master (#217) * example test commit (#182) * example test commit * example edit * ci: run test workflow on develop branch * ci: add develop branch to job triggers * ci: fix syntax issue in workflow * fix: fixed import (brainflow updated API) * build(deps): locked pylsl==1.10.5 (#187) * Experiment Class Refactor (update to #183), converting specific experiments to subclasses (#184) * First commit * Second commit * Modifications * Lol * Lol * Incorporated N170 and p300, looking good for a PR * ssvep update * Implementing subclasses instead of loose functions * fix: fixed import (brainflow updated API) * Playing around still * Fixing import errors * Adding abstractmethod decorators * Still working on the import error * Guess what's finally working * Comments and naming ticks * More comments * Live coding demonstration * ssvep adapted * Adapting Auditory Oddball * changing save_fn to self.save_fun * This maybe the last big change * utils file changed, changes work through cli Co-authored-by: Erik Bjäreholt <[email protected]> * Submodule added for gsoc * Adding pipelines for cli analysis (#202) * started pipelines function * almost working simple function equivalents of nb scripts * fix: fixed import (brainflow updated API) * sqc fixes for unicorn (#176) * Ignore pushes * Trying to create a cli * Stepping through the problem * First commit * Fixing pause in signal quality check * Fixing Signal quality check problem * fix the technical debt * Save path done for automated saving pdf * I feel amazing * Almost through * Update eegnb/cli/__main__.py Co-authored-by: Erik Bjäreholt <[email protected]> * Trying to create cli but it's being really painful * Extra word cli error * Changed example handling * Pain * Adding whole datapath * Finally fixed cli * hmm * Looking good * added hyperlink * Having some issues with detecting css and image deltetion * Just the css now * Fixed the css linking problem though it's a weird soln * Automated running, still fnames problem * Hahahah embedded images in html * Improving code * Okay now * Look at that * Almost there just the two figures now * Now * Added attrdict to do with cli error Co-authored-by: John Griffiths <[email protected]> Co-authored-by: Erik Bjäreholt <[email protected]> Co-authored-by: John Griffiths <[email protected]> * added more options for site args; improved function names; removed some redundant lines (#209) * fix subject num parsing bug * analysis report function improvements for openbci cyton and gtec unicorn devices * run exp fix * Update requirements.txt * fixes to get docs building by github action (#210) * fixes to get docs building by github action * reverted some changes * Update 01r__ssvep_viz.py Co-authored-by: John Griffiths <[email protected]> * Update README.rst small commit to test doc build workflow on this branch * removing gsoc submodule Co-authored-by: Erik Bjäreholt <[email protected]> Co-authored-by: Parv Agarwal <[email protected]> Co-authored-by: Parvfect <[email protected]> Co-authored-by: Ben Pettit <[email protected]> * update dependencies - seaborn * docs/perf: reduced the imports: `cueing` example * bug: update deprecated `plot_psd()` method - feature of new `mne` version. - instead of doing plot_psd() from the `mne.io.Raw` object, must do this: - `raw.compute_psd().plot()` - i.e., has to pass through a `spectrum` object * updated deprec. `mne` function * perf: removed importage of unused packages from example - One of them, i.e., `collections.Iterable` is even deprecated. - Must use `collections.abc.Iterable` instead now. - Resulting in faster build/user run * bugfix: `plot_conditions` - due to `sns` deprecation * bugfix: resolved `psd_welch()` deprecation (`mne`) --------- Co-authored-by: John Griffiths <[email protected]> Co-authored-by: Erik Bjäreholt <[email protected]> Co-authored-by: Parv Agarwal <[email protected]> Co-authored-by: Parvfect <[email protected]> Co-authored-by: Ben Pettit <[email protected]> Co-authored-by: Taha Morshedzadeh <[email protected]> * Updated psychopy (#215) * update psychopy for psychxr compatibility * updated n170 example to run again. * Updated doc examples for N170, P300 and SSVEP after Experiment Class Refactor (#218) * major update: merging develop to master (#217) * example test commit (#182) * example test commit * example edit * ci: run test workflow on develop branch * ci: add develop branch to job triggers * ci: fix syntax issue in workflow * fix: fixed import (brainflow updated API) * build(deps): locked pylsl==1.10.5 (#187) * Experiment Class Refactor (update to #183), converting specific experiments to subclasses (#184) * First commit * Second commit * Modifications * Lol * Lol * Incorporated N170 and p300, looking good for a PR * ssvep update * Implementing subclasses instead of loose functions * fix: fixed import (brainflow updated API) * Playing around still * Fixing import errors * Adding abstractmethod decorators * Still working on the import error * Guess what's finally working * Comments and naming ticks * More comments * Live coding demonstration * ssvep adapted * Adapting Auditory Oddball * changing save_fn to self.save_fun * This maybe the last big change * utils file changed, changes work through cli Co-authored-by: Erik Bjäreholt <[email protected]> * Submodule added for gsoc * Adding pipelines for cli analysis (#202) * started pipelines function * almost working simple function equivalents of nb scripts * fix: fixed import (brainflow updated API) * sqc fixes for unicorn (#176) * Ignore pushes * Trying to create a cli * Stepping through the problem * First commit * Fixing pause in signal quality check * Fixing Signal quality check problem * fix the technical debt * Save path done for automated saving pdf * I feel amazing * Almost through * Update eegnb/cli/__main__.py Co-authored-by: Erik Bjäreholt <[email protected]> * Trying to create cli but it's being really painful * Extra word cli error * Changed example handling * Pain * Adding whole datapath * Finally fixed cli * hmm * Looking good * added hyperlink * Having some issues with detecting css and image deltetion * Just the css now * Fixed the css linking problem though it's a weird soln * Automated running, still fnames problem * Hahahah embedded images in html * Improving code * Okay now * Look at that * Almost there just the two figures now * Now * Added attrdict to do with cli error Co-authored-by: John Griffiths <[email protected]> Co-authored-by: Erik Bjäreholt <[email protected]> Co-authored-by: John Griffiths <[email protected]> * added more options for site args; improved function names; removed some redundant lines (#209) * fix subject num parsing bug * analysis report function improvements for openbci cyton and gtec unicorn devices * run exp fix * Update requirements.txt * fixes to get docs building by github action (#210) * fixes to get docs building by github action * reverted some changes * Update 01r__ssvep_viz.py Co-authored-by: John Griffiths <[email protected]> * Update README.rst small commit to test doc build workflow on this branch * removing gsoc submodule Co-authored-by: Erik Bjäreholt <[email protected]> Co-authored-by: Parv Agarwal <[email protected]> Co-authored-by: Parvfect <[email protected]> Co-authored-by: Ben Pettit <[email protected]> * Updated doc examples * Update 00x__n170_run_experiment.py fix: typo in func param --------- Co-authored-by: John Griffiths <[email protected]> Co-authored-by: Erik Bjäreholt <[email protected]> Co-authored-by: Ben Pettit <[email protected]> Co-authored-by: Ore O <[email protected]> * fix error breaking n170 test (#223) * fixed requirements.txt so 'pip install -e .' would work on windows. (#229) * Get CI test builds working again! (#170) * Revert "Revert PR #167: "ci: fix broken CI" (#169)" This reverts commit 2d74871. * Update Makefile * fix: Update vep.py import * Update Makefile * Update vep.py * fix: typo in makefile * fix: update BaseExperiment class reference * Update Makefile * Update vep.py * Update 01r__n170_viz.py * makefile: install libnotify4 --------- Co-authored-by: Ore O <[email protected]> * fix macos build (#245) * Add devcontainer configuration for CPU environment * Fixed the python version needed * Fix plot conditions (#257) * fixed plot_conditions functoni issues * small change to viz n170 plotting example * small change to viz p300 plotting example * fixed to plotting issue * modify plot command * update example files * fix condition label bug * fix: set layout engine to fix colorbar error --------- Co-authored-by: Ore O <[email protected]> * Started the name switch (#251) * started renaming from eegnb to eegexpy * more eegnb renaming work * more eegnb switches * more relabelling * ssvep renamings * cueing eg * update for more misc files * more of the same * update * Update and rename available_notebooks.md to available_experiments.md * Add newlogo (#260) * Add files via upload * Update README.rst * added to enable eegexpy commands (#265) * started alias of main package for imports (#266) * Add multiple installation options (#263) * added newly separated analysis and streaming utils files * modified setup and requirements to allow for optional install types * updated imports in experiment examples * fixed some typos * Update requirements.txt moved click library to stim pres requirements section * Update requirements.txt Co-authored-by: Taha Morshedzadeh <[email protected]> * Update Makefile * Update docs.yml * Update Makefile this does a full build with all dependencies for EEG-ExPy --------- Co-authored-by: Ore O <[email protected]> Co-authored-by: Taha Morshedzadeh <[email protected]> * Add support for /develop in docs site (#259) * feat: update action to build doc page in a folder * test: see if it works without if clause * test: change source for docs * Initial support for VR headsets (#241) * fixed requirements.txt so 'pip install -e .' would work on windows. * added support for n170 to run on rift * altered logic to add back in jitter * simplified logic * made fullscreen again * further simplified logic * decreased font size of instructions for vr * fixed instructions display, further simplified logic, removed logs * made code easier to read * reverted board back to muse * Enable headlocking to fix jittering - cancel out any rotation and translation coming from the headset. * fixed use_vr parameter to be set to False by default for existing experiments. * fix macos build * reverted unnecessary changes and made p300 experiment display correctly * added vr support for p300 and ssvep * fix psychxr version * adding vr doc * updated vr doco --------- Co-authored-by: John Griffiths <[email protected]> * ssaep volumefadingfixes (#81) * fixed duration on instructions screen * put sound object instantiation inside trials loop. fixes sound fading issue --------- Co-authored-by: Erik Bjäreholt <[email protected]> Co-authored-by: Parv Agarwal <[email protected]> Co-authored-by: Parvfect <[email protected]> Co-authored-by: Ben Pettit <[email protected]> Co-authored-by: Ore Ogundipe <[email protected]> Co-authored-by: Taha Morshedzadeh <[email protected]> Co-authored-by: Ben Pettit <[email protected]> Co-authored-by: Taha Morshedzadeh <[email protected]>

…iments to subclasses (#184) * First commit * Second commit * Modifications * Lol * Lol * Incorporated N170 and p300, looking good for a PR * ssvep update * Implementing subclasses instead of loose functions * fix: fixed import (brainflow updated API) * Playing around still * Fixing import errors * Adding abstractmethod decorators * Still working on the import error * Guess what's finally working * Comments and naming ticks * More comments * Live coding demonstration * ssvep adapted * Adapting Auditory Oddball * changing save_fn to self.save_fun * This maybe the last big change * utils file changed, changes work through cli Co-authored-by: Erik Bjäreholt <[email protected]>

Parvfect added 9 commits May 30, 2022 18:38

First commit

1634e81

Second commit

d48093a

Modifications

f05f1c4

Lol

38993e6

Lol

ba01229

Merge branch 'parv' into experimentClassFirstRun

19ffcdc

Incorporated N170 and p300, looking good for a PR

42d3f26

ssvep update

76c5036

Implementing subclasses instead of loose functions

dd958d2

Parvfect added the enhancement New feature or request label Jun 15, 2022

Parvfect requested review from JohnGriffiths and ErikBjare June 15, 2022 12:09

This was referenced Jun 16, 2022

Run the Draft Pull Request #184 to make sure all the functionality is the same with the new structure. #186

Closed

Package Experimental parameters passing (n_trials, jitter, etc.) to Experiment super class to make it more generic. #188

Open

ErikBjare and others added 7 commits June 16, 2022 16:51

fix: fixed import (brainflow updated API)

f208e08

Playing around still

5db55f3

Fixing import errors

efb1928

Merge branch 'master' of https://github.com/NeuroTechX/eeg-notebooks …

251d04e

…into experimentSubclass2

Adding abstractmethod decorators

7436444

Still working on the import error

ee7cca9

Guess what's finally working

d73abe0

JohnGriffiths requested changes Jun 30, 2022

View reviewed changes

ErikBjare reviewed Jul 4, 2022

View reviewed changes

Comments and naming ticks

ee5ab09

Parvfect added 2 commits July 6, 2022 12:14

More comments

cb7895b

Live coding demonstration

4815d28

Parvfect marked this pull request as ready for review July 7, 2022 19:44

ssvep adapted

3ec461a

Adapting Auditory Oddball

b1000bd

This was linked to issues Jul 21, 2022

Adapt Visual SSVEP into Experimental class structure and test it #193

Closed

Adapt Auditory Oddball Experiment in the new Experimental Structure #195

Closed

Test Experimental Class Structure using EEG #196

Closed

Parvfect added 2 commits July 25, 2022 21:06

changing save_fn to self.save_fun

d9f53b2

This maybe the last big change

dc97fb5

utils file changed, changes work through cli

ef3dba0

JohnGriffiths merged commit e4e7d8a into develop Aug 10, 2022

ErikBjare reviewed Aug 11, 2022

View reviewed changes

Parvfect mentioned this pull request Jan 19, 2023

Implement version + release schedule #216

Open

15 tasks

Provide feedback

Saved searches

Use saved searches to filter your results more quickly

Experiment Class Refactor (update to #183), converting specific experiments to subclasses #184

Experiment Class Refactor (update to #183), converting specific experiments to subclasses #184

Parvfect commented Jun 15, 2022 •

edited

Loading

Parvfect commented Jun 27, 2022 •

edited

Loading

JohnGriffiths left a comment

JohnGriffiths Jun 30, 2022

JohnGriffiths Jun 30, 2022

JohnGriffiths Jun 30, 2022

Parvfect Jul 6, 2022

JohnGriffiths Jun 30, 2022

ErikBjare Jul 4, 2022

Parvfect Jul 6, 2022

JohnGriffiths Jun 30, 2022

ErikBjare left a comment

ErikBjare Jul 4, 2022

Parvfect Jul 6, 2022

ErikBjare Jul 4, 2022

Parvfect Jul 6, 2022

Parvfect commented Jul 6, 2022

JohnGriffiths commented Jul 21, 2022

Parvfect commented Jul 29, 2022

Parvfect commented Jul 29, 2022

Parvfect commented Aug 3, 2022 •

edited

Loading

Parvfect commented Aug 4, 2022

JohnGriffiths commented Aug 4, 2022

Parvfect commented Aug 5, 2022

JohnGriffiths commented Aug 9, 2022 •

edited

Loading

JohnGriffiths commented Aug 10, 2022

ErikBjare left a comment

ErikBjare Aug 11, 2022

ErikBjare Aug 11, 2022 •

edited

Loading

ErikBjare Aug 11, 2022

		prefs.hardware['audioLib'] = 'PTB'
		prefs.hardware['audioLatencyMode'] = 3

Experiment Class Refactor (update to #183), converting specific experiments to subclasses #184

Experiment Class Refactor (update to #183), converting specific experiments to subclasses #184

Conversation

Parvfect commented Jun 15, 2022 • edited Loading

Parvfect commented Jun 27, 2022 • edited Loading

JohnGriffiths 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

ErikBjare 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

Parvfect commented Jul 6, 2022

JohnGriffiths commented Jul 21, 2022

Parvfect commented Jul 29, 2022

Parvfect commented Jul 29, 2022

Parvfect commented Aug 3, 2022 • edited Loading

Parvfect commented Aug 4, 2022

For Reviewer

JohnGriffiths commented Aug 4, 2022

Parvfect commented Aug 5, 2022

JohnGriffiths commented Aug 9, 2022 • edited Loading

JohnGriffiths commented Aug 10, 2022

ErikBjare left a comment

Choose a reason for hiding this comment

Choose a reason for hiding this comment

ErikBjare Aug 11, 2022 • edited Loading

Choose a reason for hiding this comment

Choose a reason for hiding this comment

Parvfect commented Jun 15, 2022 •

edited

Loading

Parvfect commented Jun 27, 2022 •

edited

Loading

Parvfect commented Aug 3, 2022 •

edited

Loading

JohnGriffiths commented Aug 9, 2022 •

edited

Loading

ErikBjare Aug 11, 2022 •

edited

Loading