New GIDCollection print

[stinebuu]

Change the way we print GIDCollections, see #8 for details.

Printing a primitive GIDCollection without metadata will look like:

GIDCollection(metadata=None model=iaf_psc_alpha, size=5, (1..5))

Printing a layer will look like:

GIDCollection(metadata=spatial model=iaf_psc_alpha, size=25, (54..78))

Printing a sliced composite with more than 7 parts will look like:

GIDCollection(metadata=None, size=17:
              model=iaf_psc_alpha, size=3, (1..{2}..5);
              model=iaf_psc_delta, size=1, (7);
              model=iaf_psc_exp, size=1, (9);
              ..
              model=spike_detector, size=1, (19);
              model=iaf_psc_alpha, size=2, (21, 23);
              model=iaf_cond_alpha, size=5, (45..{2}..53))

[heplesser]

Nice, but I am wondering a bit about the stepping notation

 model=iaf_psc_alpha, size=3, (1..{2}..5);

Do we need the curly braces? And maybe we could make the notation even more consistent like this:

 model=iaf_psc_alpha, size=3, gids=1..2..5;
 model=iaf_psc_alpha, size=3, gids=6..8;
 model=iaf_psc_alpha, size=1, gids=9;

and similarly for non-composite.

[hakonsbm]

@heplesser My idea when adding the curly braces to the stepping notation was to separate number of steps from GIDs, because I thought writing (1..2..5) could look like it means

(GID=1 [some GIDs] GID=2 [some GIDs] GID=5)

[heplesser]

Maybe we should have a little poll on NEST User? I also thought about the Python notation 1:5:2, but that is confusing, since in real Python that is on 1 and 3, 5 is excluded.

[stinebuu]

@heplesser I like the notation of gids=6..8 for consistency. Concerning the stepping, would

model=iaf_psc_alpha, size=3, gids=1..5, step=2;

be an option?

[heplesser]

I am not totally happy with the notation, since it puts step at the same level as gids. How about

model=iaf_psc_alpha, size=3, gids=1..5 [step 2];

Should we have a discussion and vote on Monday?

[stinebuu]

@heplesser Discussion and vote sounds good!

[stinebuu]

From NEST VC: Use first, last and step:
model=iaf_psc_alpha, size=3, first=1, last=5, step=2;

Also, the first size variable from the sliced composite above makes it look like there are 17 parts of the composite, not 17 GIDs.

[stinebuu]

With 97e33f0 we will have the following

Printing a primitive GIDCollection without metadata:

GIDCollection(metadata=None, model=iaf_psc_alpha, size=5, first=1, last=5)

Printing a composite GIDCollection without any slicing:

GIDCollection(metadata=None,
              model=iaf_psc_alpha, size=5, first=1, last=5;
              model=iaf_psc_delta, size=3, first=6, last=8;
              model=iaf_psc_exp, size=1, first=9;
              model=iaf_cond_alpha, size=7, first=10, last=16;
              model=poisson_generator, size=1, first=17;
              model=spike_detector, size=3, first=18, last=20)

Printing a sliced composite with more than 7 parts:

GIDCollection(metadata=None,
              model=iaf_psc_alpha, size=3, first=1, last=5, step=2;
              model=iaf_psc_delta, size=1, first=7, last=7;
              model=iaf_psc_exp, size=1, first=9, last=9;
              ..
              model=spike_detector, size=1, first=19, last=19;
              model=iaf_psc_alpha, size=2, first=21, last=23, step=2;
              model=iaf_cond_alpha, size=5, first=45, last=53, step=2)

Printing a layer:

GIDCollection(metadata=spatial, model=iaf_psc_alpha, size=25, first=54, last=78)

@heplesser @hakonsbm what do you think?

[stinebuu]

@heplesser @hakonsbm as discussed, I have now removed the partitioning when a composite have more than 7 parts.

[heplesser]

@stinebuu So what happens now if you print a GC with more than seven parts?

[stinebuu]

@heplesser You are shown all the parts of the GIDCollection, like this, where we have a composite gc which is sliced and consists of 8 primitives:

GIDCollection(metadata=None,
              model=iaf_psc_alpha, size=3, first=1, last=5, step=2;
              model=iaf_psc_delta, size=1, first=7, last=7;
              model=iaf_psc_exp, size=1, first=9, last=9;
              model=iaf_cond_alpha, size=3, first=11, last=15, step=2;
              model=poisson_generator, size=1, first=17, last=17;
              model=spike_detector, size=1, first=19, last=19;
              model=iaf_psc_alpha, size=2, first=21, last=23, step=2;
              model=iaf_cond_alpha, size=5, first=45, last=53, step=2)

[heplesser]

OK.

…

On 16 Apr 2019, at 13:19, Stine Brekke Vennemo ***@***.***> wrote: @heplesser You are shown all the parts of the GIDCollection, like this, where we have a composite gc which is sliced and consists of 8 primitives: GIDCollection(metadata=None, model=iaf_psc_alpha, size=3, first=1, last=5, step=2; model=iaf_psc_delta, size=1, first=7, last=7; model=iaf_psc_exp, size=1, first=9, last=9; model=iaf_cond_alpha, size=3, first=11, last=15, step=2; model=poisson_generator, size=1, first=17, last=17; model=spike_detector, size=1, first=19, last=19; model=iaf_psc_alpha, size=2, first=21, last=23, step=2; model=iaf_cond_alpha, size=5, first=45, last=53, step=2) — You are receiving this because you were mentioned. Reply to this email directly, view it on GitHub, or mute the thread.

-- Prof. Dr. Hans Ekkehard Plesser Head, Data Science Section Faculty of Science and Technology Norwegian University of Life Sciences PO Box 5003, 1432 Aas, Norway Phone +47 6723 1560 Email [email protected] Home http://arken.nmbu.no/~plesser