IMI Neuro issueshttps://git.rwth-aachen.de/fabian.schlebusch/imi-neuro/-/issues2021-04-29T08:11:30+02:00https://git.rwth-aachen.de/fabian.schlebusch/imi-neuro/-/issues/13Documentation2021-04-29T08:11:30+02:00Aidan NickersonDocumentationHi Fabian,
I'm trying to add code for the wavelet filter, and add adapter for our datasets - but there is little in the way of documentation.
There are also several notebooks at the root level, but its not clear what these do and they ...Hi Fabian,
I'm trying to add code for the wavelet filter, and add adapter for our datasets - but there is little in the way of documentation.
There are also several notebooks at the root level, but its not clear what these do and they seem to not work.
Adding a readme might be useful :)https://git.rwth-aachen.de/fabian.schlebusch/imi-neuro/-/issues/11Identify and move reusable code from neo spike importer2021-02-23T15:29:12+01:00Fabian SchlebuschIdentify and move reusable code from neo spike importerThere are functions in the neo spike importer which might be reusable for other importers. This includes the preparation of the channels and other neo objects. After we did that, it might make sense to revisit the dapsys importer and see...There are functions in the neo spike importer which might be reusable for other importers. This includes the preparation of the channels and other neo objects. After we did that, it might make sense to revisit the dapsys importer and see if we can make use of these functions.Frederic KehreinFabian SchlebuschFrederic Kehreinhttps://git.rwth-aachen.de/fabian.schlebusch/imi-neuro/-/issues/8Bug: Adaptive Spike Count Length always 1 bigger than the defined number, pro...2020-12-01T14:16:12+01:00DaminBug: Adaptive Spike Count Length always 1 bigger than the defined number, prob related to Issue 4`AdaptiveSpikeCount(actpots = recording.actpots, timeframe = 100, num_splits = 8)`
Expected: `[0.,0.,0.,0.,0.,0.,0.,0.]`
Got: `[0.,0.,0.,0.,0.,0.,0.,0.,0.]``AdaptiveSpikeCount(actpots = recording.actpots, timeframe = 100, num_splits = 8)`
Expected: `[0.,0.,0.,0.,0.,0.,0.,0.]`
Got: `[0.,0.,0.,0.,0.,0.,0.,0.,0.]`https://git.rwth-aachen.de/fabian.schlebusch/imi-neuro/-/issues/7Feature: GUI Import Window saving state2020-11-20T12:38:22+01:00DaminFeature: GUI Import Window saving stateWhen importing data a lot of things can go wrong (Running out of RAM etc.) which means that often this process has to be startet multiple times in a row.
Always having to go through the interface makes this process longer than needed, e...When importing data a lot of things can go wrong (Running out of RAM etc.) which means that often this process has to be startet multiple times in a row.
Always having to go through the interface makes this process longer than needed, expecially if the set values/channels were correct or almost correct.
The possibility to have the last input values show up in the interface if only a few changes are needed would be great. Obvs. you could always comment out the GUI if you don't want to change the values but this feature might make it more user friendly.https://git.rwth-aachen.de/fabian.schlebusch/imi-neuro/-/issues/6Feature: ActionPotential provides function from_dataframe, but no to_dataframe2020-11-20T12:38:11+01:00DaminFeature: ActionPotential provides function from_dataframe, but no to_dataframeCurrently iterating over all APs extracting each feature + filtering for empty values in SpikeCount (see [Issue 4](https://git.rwth-aachen.de/fabian.schlebusch/imi-neuro/-/issues/4))
Might be a useful function for future users.Currently iterating over all APs extracting each feature + filtering for empty values in SpikeCount (see [Issue 4](https://git.rwth-aachen.de/fabian.schlebusch/imi-neuro/-/issues/4))
Might be a useful function for future users.https://git.rwth-aachen.de/fabian.schlebusch/imi-neuro/-/issues/5Bug(?): SpikeCount needs to be converted to Float, before being converted to Int2020-11-20T12:37:33+01:00DaminBug(?): SpikeCount needs to be converted to Float, before being converted to IntCurrently the feature spike_count is not an integer, eventhough we only expect integer values.
When casting it to an numpy array with type int we get:
`invalid literal for int() with base 10: '0.'`
To avoid this it needs to be castet t...Currently the feature spike_count is not an integer, eventhough we only expect integer values.
When casting it to an numpy array with type int we get:
`invalid literal for int() with base 10: '0.'`
To avoid this it needs to be castet to float before, which results in something like this:
`***.astype(float).astype(int)`
We might speed up the preprocessing + save space and ram if saved as integers.https://git.rwth-aachen.de/fabian.schlebusch/imi-neuro/-/issues/4Bug: SpikeCount has additional, empty entries2020-12-01T14:15:53+01:00DaminBug: SpikeCount has additional, empty entriesFor some APs the spike_count has more entires than the number of defined intervalls which are empty (= ''). Reproducable with e.g. Spike_Animal+Mech_01 data from 1100sec on for the 3. AP (can provide more infomration if needed)
Expected...For some APs the spike_count has more entires than the number of defined intervalls which are empty (= ''). Reproducable with e.g. Spike_Animal+Mech_01 data from 1100sec on for the 3. AP (can provide more infomration if needed)
Expected:
`'spike_count': ['0.', '0.', '0.', '0.', '0.', '0.', '0.', '0.', '0.', '0.', '0.', '0.', '1.', '0.', '29.', '0.']`
Got:
`'spike_count': ['', '0.', '0.', '0.', '0.', '0.', '0.', '0.', '0.', '0.', '0.', '0.', '0.', '1.', '0.', '29.', '0.']`
Current Workaround:
`numpy.delete(ap_np, ap_np == '')`https://git.rwth-aachen.de/fabian.schlebusch/imi-neuro/-/issues/2Model a C-Fibre using classes2020-09-02T10:47:33+02:00Fabian SchlebuschModel a C-Fibre using classesWe need a 1:1 model for a C-fibre (candidate). The class should contain e.g.:
- the AP template we use for correlation
- a list of all APs that belong to this fibre
- information about the fibre: which type, conduction speed etc.We need a 1:1 model for a C-fibre (candidate). The class should contain e.g.:
- the AP template we use for correlation
- a list of all APs that belong to this fibre
- information about the fibre: which type, conduction speed etc.Fabian SchlebuschFabian Schlebusch