| Re: frustrated by hdf_sd_adddata [message #24182] |
Thu, 15 March 2001 10:52  |
Liam E. Gumley
Messages: 378
Registered: January 2000
|
Senior Member |
|
|
George McCabe wrote:
> attempting to replace a subset of existing data in an HDF SD, using
> idl procedure hdf_sd_adddata, receive the following error which i can
> not readily diagnose -
> % HDF_SD_ADDDATA: Unable to write the specified HDF-SD slice.
>
> hdf opens, sd starts /RDWR ok, sd selected ok
> i have veryfied the size and order of the data variable, count and
> start keyword variables.
George,
Does the following example work on your system?
hdfid = hdf_sd_start('test.hdf')
varid = hdf_sd_create(hdfid, 'data', [256, 256], /float)
hdf_sd_adddata, varid, dist(256)
hdf_sd_endaccess, varid
hdf_sd_end, hdfid
If it does, the HDF interface on you platform is working. You might be
having a problem with the definition of the START and COUNT keywords.
The following is from the IDL 5.3 online documentation:
---begin quote---
START
Set this keyword to a vector that contains the starting position for the
data. The default position is [0, 0, ..., 0].
COUNT
Set this keyword to a vector of counts (i.e., the number of items) to be
written in each dimension. The default is to write all available data.
STRIDE
Set this keyword to a vector that contains the strides, or sampling
intervals, between accessed values of the NetCDF variable. The default
stride vector is that for a contiguous write: [0, 0, ..., 0].
---end quote---
I'd try and reduce the code to the smallest number of lines which
reproduces the error. This process often makes the source of the error
obvious.
Cheers,
Liam.
http://cimss.ssec.wisc.edu/~gumley
|
|
|
|
|
|
|
|
|
|
| Re: frustrated by hdf_sd_adddata [message #24198 is a reply to message #24186] |
Thu, 15 March 2001 15:47  |
John-David T. Smith
Messages: 384
Registered: January 2000
|
Senior Member |
|
|
"Pavel A. Romashkin" wrote:
>
> George McCabe wrote:
>>
>> this is the microsoft service model right? software modules that fail
>> require system restart.
>>
>> sorry to be negative
>
> I am really happy that at least there is *something* that helps. We
> today take a working computer/software for granted. Given how complex
> the processors, OSs and applications are these days, I am glad and
> surprised they work at all. And a restart takes just a few seconds. What
> are we complaining about? Our cars stall and need restarts, they break
> and require engine rebuilds, and we don't demand perpetual engines for them.
vodka% uptime
6:41pm up 138 days, 21:02, 3 users, load average: 0.02, 0.09, 0.05
Unless you do weird hardware things to it, Linux *never* crashes. It's
frightening what a little protected memory can do for you.
;)
jd
|
|
|
|
| Re: frustrated by hdf_sd_adddata [message #24199 is a reply to message #24187] |
Thu, 15 March 2001 11:00 |
davidf
Messages: 2866
Registered: September 1996
|
Senior Member |
|
|
Pavel A. Romashkin (pavel.romashkin@noaa.gov) writes:
> They recommend to always try to reproduce a problem after a complete IDL restart.
They don't just recommend it. If they are any good they
require it. When I was a technical support engineer we
found that 80% of the calls could be solved just by this
one simple procedure.
But I find I still forget to use it 75% of the time when
I'm trying to figure out my own stupid mistakes. :-(
Cheers,
David
--
David Fanning, Ph.D.
Fanning Software Consulting
Phone: 970-221-0438 E-Mail: davidf@dfanning.com
Coyote's Guide to IDL Programming: http://www.dfanning.com/
Toll-Free IDL Book Orders: 1-888-461-0155
|
|
|
|
| Re: frustrated by hdf_sd_adddata [message #24200 is a reply to message #24187] |
Thu, 15 March 2001 08:18 |
George McCabe
Messages: 23
Registered: June 1997
|
Junior Member |
|
|
<!doctype html public "-//w3c//dtd html 4.0 transitional//en">
<html>
james,
<p>thanks for you many suggestions.
<p>can you tell me what do you mean by 'corrupt'. though i think that i
understand the use of idl's hdf procedures, my general experience is limited.
i am curious about occurances of hdf file corruption.
<p>this discussion is perhaps no longer relevant to my particular problem+solution.
today, under a new idl session writing data to the same HDF file with
the same code it works, so i am wondering about software interface errors,
buffer problems, etc. maybe i'll just pretend it never happened.
<p>thank you again,
<br>george
<br>
<p>James Kuyper wrote:
<blockquote TYPE=CITE>George McCabe wrote:
<br>>
<br>> hello,
<br>>
<br>> attempting to replace a subset of existing data in an HDF SD, using
<br>> idl procedure hdf_sd_adddata, receive the following error which i
can
<br>> not readily diagnose -
<br>> % HDF_SD_ADDDATA: Unable to write the specified HDF-SD slice.
<br>>
<br>> hdf opens, sd starts /RDWR ok, sd selected ok
<br>> i have veryfied the size and order of the data variable, count and
<br>> start keyword variables.
<br>>
<br>> help is appreciated,
<br>> george
<p>Some things to check:
<p>You might not have access. What are the file permissions, owner, and
<br>group membership?
<p>The file may be corrupt. Can you view the HDF file with ncdump, or
<br>similar utilities? Can you open it read-only, and view its contents
<br>using HDF_SD_GETDATA?
<p>Is this new code, or code that has already worked on a different file,
<br>but is failing with this one?
<p>You imply that you've already collected the following information, and
<br>that it checked out OK, but maybe you've misinterpreted it - could
you
<br>tell us:
<p>What values does HDF_SD_GETINFO give you for DIMS, HDF_TYPE, and TYPE?
<p>What values are you giving HDF_SD_ADDDATA for START, COUNT, and STRIDE?
<p>What do you get when you type
<p> HELP,mydata
<p>where 'mydata' represents the name of the field you're passing as the
<br>data argument to HDF_SD_ADDDATA?</blockquote>
<pre>--
George McCabe (CUofA)
NASA / Goddard Space Flight Center / Code 693
Building 2, Room 151 / Greenbelt, MD 20771
voice: (301)286-8283, fax: (301)286-0212
george.mccabe@gsfc.nasa.gov</pre>
</html>
|
|
|
|
|
|
| Re: frustrated by hdf_sd_adddata [message #24209 is a reply to message #24200] |
Thu, 15 March 2001 06:13 |
George McCabe
Messages: 23
Registered: June 1997
|
Junior Member |
|
|
<!doctype html public "-//w3c//dtd html 4.0 transitional//en">
<html>
i checked that david. without write permission i would not be able to hdf_sd_start
as in the following,
<p>IDL> sd_id=HDF_SD_START(f,/RDWR)
<br>% HDF_SD_START: Unable to start the HDF-SD interface.
<p>which i am able to do.
<p>thank you,
<br>george
<p>David Fanning wrote:
<blockquote TYPE=CITE>George McCabe (george.mccabe@gsfc.nasa.gov) writes:
<p>> attempting to replace a subset of existing data in an HDF SD, using
idl
<br>> procedure hdf_sd_adddata, receive the following error which i can
not
<br>> readily diagnose -
<br>> % HDF_SD_ADDDATA: Unable to write the specified HDF-SD slice.
<br>>
<br>> hdf opens, sd starts /RDWR ok, sd selected ok
<br>> i have veryfied the size and order of the data variable, count and
start
<br>> keyword variables.
<br>>
<br>> help is appreciated,
<p>No file writing permissions?
<p>Cheers,
<p>David
<br>--
<br>David Fanning, Ph.D.
<br>Fanning Software Consulting
<br>Phone: 970-221-0438 E-Mail: davidf@dfanning.com
<br>Coyote's Guide to IDL Programming: <a href="http://www.dfanning.com/">http://www.dfanning.com/</a>
<br>Toll-Free IDL Book Orders: 1-888-461-0155</blockquote>
<pre>--
George McCabe (CUofA)
NASA / Goddard Space Flight Center / Code 693
Building 2, Room 151 / Greenbelt, MD 20771
voice: (301)286-8283, fax: (301)286-0212
george.mccabe@gsfc.nasa.gov</pre>
</html>
|
|
|
|
| Re: frustrated by hdf_sd_adddata [message #24210 is a reply to message #24209] |
Thu, 15 March 2001 07:50 |
James Kuyper
Messages: 425
Registered: March 2000
|
Senior Member |
|
|
George McCabe wrote:
>
> hello,
>
> attempting to replace a subset of existing data in an HDF SD, using
> idl procedure hdf_sd_adddata, receive the following error which i can
> not readily diagnose -
> % HDF_SD_ADDDATA: Unable to write the specified HDF-SD slice.
>
> hdf opens, sd starts /RDWR ok, sd selected ok
> i have veryfied the size and order of the data variable, count and
> start keyword variables.
>
> help is appreciated,
> george
Some things to check:
You might not have access. What are the file permissions, owner, and
group membership?
The file may be corrupt. Can you view the HDF file with ncdump, or
similar utilities? Can you open it read-only, and view its contents
using HDF_SD_GETDATA?
Is this new code, or code that has already worked on a different file,
but is failing with this one?
You imply that you've already collected the following information, and
that it checked out OK, but maybe you've misinterpreted it - could you
tell us:
What values does HDF_SD_GETINFO give you for DIMS, HDF_TYPE, and TYPE?
What values are you giving HDF_SD_ADDDATA for START, COUNT, and STRIDE?
What do you get when you type
HELP,mydata
where 'mydata' represents the name of the field you're passing as the
data argument to HDF_SD_ADDDATA?
|
|
|
|
| Re: frustrated by hdf_sd_adddata [message #24211 is a reply to message #24200] |
Thu, 15 March 2001 09:45 |
James Kuyper
Messages: 425
Registered: March 2000
|
Senior Member |
|
|
George McCabe wrote:
>
> james,
>
> thanks for you many suggestions.
>
> can you tell me what do you mean by 'corrupt'. though i think that i
> understand the use of idl's hdf procedures, my general experience is
> limited. i am curious about occurances of hdf file corruption.
When I talk about file corruption, I'm not talking about anything
particularly exotic. HDF expects the input file to have data is a
certain format. If the file is damaged, the data might not be in the
right format any more. Ways to damage a file include using non-HDF
facilities to modify the file, doing file transfers with no error
checking, using text mode ftp, or physical damage to your storage
medium.
There are probably ways to damage an HDF file by passing bad arguments
to HDF functions, but I've never bothered testing to see which kinds of
bad arguments HDF actually checks for, and which ones it lets through.
|
|
|
|
| Re: frustrated by hdf_sd_adddata [message #24218 is a reply to message #24209] |
Wed, 14 March 2001 16:33 |
davidf
Messages: 2866
Registered: September 1996
|
Senior Member |
|
|
George McCabe (george.mccabe@gsfc.nasa.gov) writes:
> attempting to replace a subset of existing data in an HDF SD, using idl
> procedure hdf_sd_adddata, receive the following error which i can not
> readily diagnose -
> % HDF_SD_ADDDATA: Unable to write the specified HDF-SD slice.
>
> hdf opens, sd starts /RDWR ok, sd selected ok
> i have veryfied the size and order of the data variable, count and start
> keyword variables.
>
> help is appreciated,
No file writing permissions?
Cheers,
David
--
David Fanning, Ph.D.
Fanning Software Consulting
Phone: 970-221-0438 E-Mail: davidf@dfanning.com
Coyote's Guide to IDL Programming: http://www.dfanning.com/
Toll-Free IDL Book Orders: 1-888-461-0155
|
|
|
|
comp.lang.idl-pvwave archive
Members
Search
Help
Login
Home
