clear;//// read input files//q=3D400;disp('Loading data');filename =3D 'G:\=
PROJECTS\Besson\AURA2008\Sparky\TestData\test001\zoom001.txt';x =3D read(fi=
lename, q, 6);y1 =3D 10*x(:,5);
The argument list for "read" is "filename,m,n,[format]" and "filename,m,n,k=
,format"
In my simple example, I'm reading only the first 400 lines and the first 6 =
columns.  Then I create a temporary variable y1, that is only the data from=
 column 5 scaled up by 10.
I could read any sub-matrix from the file this way.  These data files weren=
't terribly large but there were several of them.  Later in the script I ha=
ve code to modify "filename"as needed to read all the files in the sequence=
..
To write back out y1 I would use "save" for the binary formatted file (fast=
) or "write" for the ASCII formatted version (slow).

I didn't specify a format but you could use this string to help include or =
exclude data from a CSV that has text mixed in.=20
To read the entire CSV file, assuming there were no lines of text, just num=
bers and commas, it would be x=3Dread(filename,-1,-1);
Like I said, "read" is slow.  Here I'm only using the first 400 data points=
, but I had a bunch of files.  In another example, I had 10,000,000 data po=
ints and that just takes a while
for the first read.=20
You could add code to your M file using the "fileinfo" command where you fi=
rst check to see if the binary version of your data file exists, if not, re=
ad the CSV,create the binary and then compute.  Then the next time you run =
the script, it would discover the binary version and could read that instea=
d.
Rob

> Date: Wed, 19 Sep 2012 20:30:52 +0100
> From: lists@edeca.net
> To: piclist@mit.edu
> Subject: Re: [EE] FFT for Windows
>=20
> On 19/09/2012 20:20, Robert Young wrote:
> > I've had mostly good luck with SciLab but its read of CSV files is pret=
ty slow.
> >
> > One thing I do to speed it up a bit is tell SciLab which columns I'm in=
terested in reading.  Then the very next command after "read" is a write to=
 push the data back out into a SciLab data file.  Those read and write much=
 faster.
>=20
> Would you mind providing a sample of the commands you would run to=20
> ingest the data and save out as a SciLab file?  That might give me a=20
> useful headstart.
>=20
> I can happily massage the data into a sensible format before ingesting=20
> but it would be quicker to work with the CSV straight from the scope.
>=20
> David
> --=20
> http://www.piclist.com PIC/SX FAQ & list archive
> View/change your membership options at
> http://mailman.mit.edu/mailman/listinfo/piclist
 		 	   		 =20
--=20
http://www.piclist.com PIC/SX FAQ & list archive
View/change your membership options at
http://mailman.mit.edu/mailman/listinfo/piclist
.