On 10/01/2015 06:12 AM, Robert Hanson wrote:
> Then map.array("id") is perfect for you. It turns that map into an actual
> array (temporarily), and you can then use WHERE and WHEREIN on that next
> level of maps. After you do that, using array("id") again returns your
> top-level map. The result was identical to what you requested.
>
I couldn't test it yet (see below) but I fear the memory footprint might
be quite high if data has to be copied for it.
The memory footprint of multi-level data constructs in Jmol generally
seems to be quite high. If I load for example about 15 megabyte of
annotation data for about 200,000 atoms, calculate coordinates, add
bonds and then actually load the atoms about 900 megabyte of memory seem
to be needed. Saved on disk with 'write var...' it takes about 300
megabytes (and about 200 megabytes after removing whitespace).
In Perl my rule of thumb is a factor of 10-20 to load data into a
multi-level data structure. The numbers above sound like a factor of 40
in Jmol.
But maybe this is related also to the observation I mentioned today in
the thread "Jmol reads a structure file from a pipe or system input":
"In the application there are shown two numbers about the memory used by
Jmol at the bottom right. The amount of memory used increases about
20-30 megabytes for each command I enter into the script console,
seemingly independent of the command itself."
> I think what you are missing is that Jmol is treating an array or array of
> arrays totally differently from a map. The array is passed through, and
> WHERE or WHEREIN operate directly on the map itself, no matter how many
> levels deep it is in the array. So it makes no sense to refer to the array
> itself and its subarrays as the "first level." The first map that is found
> -- that is the level for WHERE; if the values in that map are also maps,
> those are what are tested for WHEREIN.
>
I don't know anything about the Jmol's internals, so what I get here is
that it would be more complicated than I thought to implement my proposal.
> It's not more complicated than that.
>
> snpInfo = {"rs1229984": [resno: 48,
> from: "R",
> to: "H"],
> "rs1041969": [resno: 57,
> from: "N",
> to: "K"]
> };
>
> print "=== Case 1 ====";
> print snpInfo.array("id").select("id,from,to WHERE
> to='H'").array("id").format("JSON");
>
> print "=== Case 2 ====";
> print snpInfo.array("id").select("(resno) WHERE to='H'").format("JSON");
>
> === Case 1 ====
> { "rs1229984": { "to": "H","from": "R" } }
> === Case 2 ====
> [ 48 ]
>
This example doesn't work for me in Jmol 14.3.16_2015.09.28:
------------------------------------------
$ print snpInfo.array("id").select("id,from,to WHERE
to='H'").array("id").format("JSON");
script ERROR: invalid argument
----
print snpInfo . array ( "id" ) . select ( "id,from,to WHERE
to='H'" ) . array ( "id" ) . format ( "JSON" <<<<)
$ print snpInfo.array("id").select("(resno) WHERE to='H'").format("JSON");
script ERROR: invalid argument
----
print snpInfo . array ( "id" ) . select ( "(resno) WHERE
to='H'" ) . format ( "JSON" <<<<)
-------------------------------------------
Regards,
Rolf
--
Rolf Huehne
Postdoc
Leibniz Institute for Age Research - Fritz Lipmann Institute (FLI)
Beutenbergstrasse 11
07745 Jena, Germany
Phone: +49 3641 65 6205
Fax: +49 3641 65 6210
E-Mail: [email protected]
Website: http://www.fli-leibniz.de
Scientific Director: Prof. Dr. K. Lenhard Rudolph
Head of Administration: Dr. Daniele Barthel
Chairman of Board of Trustees: Dennys Klein
VAT No: DE 153 925 464
Register of Associations: No. 230296, Amtsgericht Jena
Tax Number: 162/141/08228
------------------------------------------------------------------------------
_______________________________________________
Jmol-users mailing list
[email protected]
https://lists.sourceforge.net/lists/listinfo/jmol-users
