Allow select population, GIDs, or random for analysis

[antonelepfl]

So the current implementation looks something like:

The idea would be to select some sub-population, specify some GIDs, specify a range of GIDs, or just random cells that will be used for all the different analysis.

Some mockups of how this could like:

@mmigliore Do you think this are good or what suggestion would you propose?

[mmigliore]

Hello, the first mockup looks ok but the list of GIDs should be selected independently for histogram/raster and voltage. This because one can usually pick a much larger set for raster and histogram and a smaller number of cells to show the voltage.

On Wed, Oct 16, 2019 at 2:37 PM Stefano Antonel notifications@github.com wrote:

So the current implementation looks something like:

[image: Screenshot 2019-10-16 at 14 36 21] https://user-images.githubusercontent.com/24406041/66919833-6248c000-f022-11e9-8903-17c36642866d.png

The idea would be to select some sub-population, specify some GIDs, specify a range of GIDs, or just random cells that will be used for all the different analysis.

Some mockups of how this could like: [image: Screenshot 2019-10-16 at 14 33 57] https://user-images.githubusercontent.com/24406041/66919652-0847fa80-f022-11e9-8ee0-7ed0ca9fdae2.png [image: Screenshot 2019-10-16 at 14 33 50] https://user-images.githubusercontent.com/24406041/66919653-08e09100-f022-11e9-9e1c-ec1d2d6e1d88.png [image: Screenshot 2019-10-16 at 14 36 31] https://user-images.githubusercontent.com/24406041/66919848-6c6abe80-f022-11e9-96b2-b4fd30b3fb52.png

@mmigliore https://github.com/mmigliore Do you think this are good or what suggestion would you propose?

— You are receiving this because you were mentioned. Reply to this email directly, view it on GitHub https://github.com/cnr-ibf-pa/hbp-bsp-issues/issues/482?email_source=notifications&email_token=AGZUP3YKZTSSWIQDMGASRH3QO4DKDA5CNFSM4JBK5KLKYY3PNVWWK3TUL52HS4DFUVEXG43VMWVGG33NNVSW45C7NFSM4HSEZEXA, or unsubscribe https://github.com/notifications/unsubscribe-auth/AGZUP37DXO2YFPVLQBHHUUTQO4DKDANCNFSM4JBK5KLA .

[antonelepfl]

Thank you Michele,

I'm thinking in several options that meet your requirements: 1) You run an analysis with Raster and Histogram and select a population, in this case the voltage trace will be disabled due population may be too big. Then if you see that the network is behaving well, you run a second analysis selecting some random GIDs or specific ones and the voltage will be available.

2) We can separate the selection so it in the way that you can specify for each type of plot, which population is going to be used:

The only issue with 2) is that if you select random cells they may be different for each plot but if you want to really do a deep analysis, you can enter specific GIDs

[mmigliore]

Yes, that I think that it would work.

On Wed, Oct 16, 2019 at 5:07 PM Stefano Antonel notifications@github.com wrote:

Thank you Michele,

I'm thinking in several options that meet your requirements:

1.

You run an analysis with Raster and Histogram and select a population, in this case the voltage trace will be disabled due population may be too big. Then if you see that the network is behaving well, you run a second analysis selecting some random GIDs or specific ones and the voltage will be available. 2.

We can separate the selection so it in the way that you can specify for each type of plot, which population is going to be used:

[image: Screenshot 2019-10-16 at 16 52 13] https://user-images.githubusercontent.com/24406041/66930821-76e28380-f035-11e9-8dad-1e038fedcba3.png

[image: Screenshot 2019-10-16 at 16 52 19] https://user-images.githubusercontent.com/24406041/66930824-76e28380-f035-11e9-8676-3f40e5098572.png

[image: Screenshot 2019-10-16 at 16 52 26] https://user-images.githubusercontent.com/24406041/66930825-76e28380-f035-11e9-8685-9ea98a7fcf61.png

The only issue with 2) is that if you select random cells they may be different for each plot but if you want to really do a deep analysis, you can enter specific GIDs

— You are receiving this because you were mentioned. Reply to this email directly, view it on GitHub https://github.com/cnr-ibf-pa/hbp-bsp-issues/issues/482?email_source=notifications&email_token=AGZUP34XCP5AWEFTGNU6XCDQO4U4VA5CNFSM4JBK5KLKYY3PNVWWK3TUL52HS4DFVREXG43VMVBW63LNMVXHJKTDN5WW2ZLOORPWSZGOEBM23WA#issuecomment-542748120, or unsubscribe https://github.com/notifications/unsubscribe-auth/AGZUP3776IJRBACLTK2VPVTQO4U4VANCNFSM4JBK5KLA .

[antonelepfl]

Which one would work? option 1 or 2

[mmigliore]

Sorry, I would say option 2

On Wed, Oct 16, 2019 at 5:13 PM Stefano Antonel notifications@github.com wrote:

Which one would work? option 1 or 2

— You are receiving this because you were mentioned. Reply to this email directly, view it on GitHub https://github.com/cnr-ibf-pa/hbp-bsp-issues/issues/482?email_source=notifications&email_token=AGZUP3YUARRNSP5CWWLRXHTQO4VSXA5CNFSM4JBK5KLKYY3PNVWWK3TUL52HS4DFVREXG43VMVBW63LNMVXHJKTDN5WW2ZLOORPWSZGOEBM3RXY#issuecomment-542750943, or unsubscribe https://github.com/notifications/unsubscribe-auth/AGZUP3YGIY64VFLVPAPP5J3QO4VSXANCNFSM4JBK5KLA .

[antonelepfl]

I have implemented this on sim-ui DEV

[alex4200]

I have three remarks

• Maybe it makes more sense to set the default value for the number of traces to 5 instead of 1?
• (LowPrio) I noticed that when you choose GID e.g. '11-13' then this means only GID's 11 and 12 are used (and not 13). This is consistent with the python list implementation, but might confuse other users.
• When selecting random cell, you can only have a maximum of 80 (for PST and Spike Raster Plot). Why the limit to 80 if the 'Whole' plot has much more data points (like 1000)?

[antonelepfl]

I implemented this already

[alex4200]

Seems to work, closing ticket