[Thread Prev][Thread Next][Index]
Re: non COARDS-compliant via DODS
Hi Andrew,
Just an fyi mention of an alternative approach to the use of single
quotes in issue "2":
"2) Unfortunately, Ferret requires that variables with "unusual"
characters
be quoted with single quotes."
Since you are using an "init" script to remedy inadequacies of the
self-describing metadata in the the data set you can at the same time
rename the "unusual" variable names. In the init script this would be a
Ferret command line like
SET VARIABLE/NAME=new_name 'unusual_name'
Thereafter refer to new_name instead of 'unusual_name' in the XML
description of the data set.
This solution requires the use of Ferret V5.33 beta in LAS -- see
http://ferret.wrc.noaa.gov/Ferret/Downloads/beta_releases.html
- steve
==============================
Joe Sirott wrote:
> Andrew,
>
> Your first problem (out-of-range errors) is due to the fact that there
> is no metadata in your HDF files for the X (longitude) axis indicating
> the units for the axis. Ferret also needs another netCDF attribute for
> the axis named "modulo"; this is detailed in the COARDS conventions.
>
> The solution is to configure LAS with the <init_script> tag. I've
> attached a script that will work for your dataset. The steps to use
> this are:
>
> 1) configure Ferret to use custom directories (if you haven't already)
>
> 2) Save the attached script init_modulo_x.jnl in:
> lasxml/server/custom.
> 3) Add/change the following:
> <datasets><d4733_restart1_a_nc_dods_zip name="DODS HDF test" ...>
> <properties>
> <ferret><init_script>init_modulo_x</init_script>
> </ferret>
> </properties>
> 4) Regenerate the server with genLas2.pl (or make).
>
> You problem with bizarre variable names can be solved as follows:
>
> 1) For variables that have names that can't be translated into XML
> tags, you can create your own XML tag name and use the url attribute;
> for instance:
> <variables>
> <barotropic_v_velocity name="barotropic_v_velocity"
> url="http://www.nerc-essc.ac.uk/cgi-bin/nph-dods/OCCAM/d4733.restart1_a.hdf.gz#barotropic%20v%20velocity";>
>
> ...
> 2) Unfortunately, Ferret requires that variables with "unusual"
> characters be quoted with single quotes. This can be fixed by
> installing the attached file 'custom.pl' in the lasxml/server/custom
> directory.
>
> Andrew Woolf wrote:
>
>> Thanks Joe and others with help on my question the other day
>> (addXml.pl).
>>
>> I'm trying to set up LAS for our data, which is currently in non
>> COARDS-compliant HDF format. I could convert all of these, but there
>> is a
>> huge amount of data (~150GB compressed), so I'd prefer if I could
>> get it
>> all working as is.
>>
>> In order to do this I've currently got a DODS server running and am
>> trying
>> to read the data in LAS via DODS.
>>
>> Of course I need to hand edit las.xml for the configuration.
>>
>> *So*: my data has 2-d variables (lat,lon) and 3-d (lat,lon,depth). I
>> set up
>> las.xml defining the grids for these variables with the axes in the
>> correct order, but it produces an error when asking ferret to do the
>> plotting.
>>
>> An example is the DODS file
>> http://www.nerc-essc.ac.uk/
>> gi-bin/nph-dods/OCCAM/d4733.restart1_a.hdf.gz
>>
>> An ncdump shows salinity as:
>>
>> float SALINITY(LATITUDE_T, LONGITUDE_T, DEPTH) ;
>> SALINITY:long_name = "Salinity" ;
>> SALINITY:units = "(PSU-35.)/1000." ;
>> SALINITY:_FillValue = 0.0099999998f ;
>> SALINITY:LEVEL = 0 ;
>> SALINITY:T_GRID = 1 ;
>>
>> Coordinate vars are:
>> * LATITUDE_T (577 pts equally spaced between -78.125 and 65.875)
>> * LONGITUDE_T (1440 equally spaced between 0.125 and 359.875)
>> * DEPTH (lets say 1:36 to keep things simple)
>>
>> (Ignore all variables except salinity, because is another problem
>> with
>> underscores in other variable names being translated to spaces....).
>>
>> For a trial I set this up in las.xml as follows:
>>
>> <d4733_restart1_a_nc_dods_zip name="DODS HDF test"
>> url="http://www.nerc-essc.ac.uk/cgi-bin/nph-dods/OCCAM/d4733.restart1_a.hdf.gz";
>>
>> doc="">
>> <variables>
>> ...
>> <salinity units="PSU" name="Salinity">
>> <link
>> match="/lasdata/grids/OCCAM_LATITUDE_T_LONGITUDE_T_DEPTH_grid"/>
>> </salinity>
>> ...
>> </variables
>> </d4733_restart1_a_nc_dods_zip>
>>
>> ...
>>
>> <grids>
>> ...
>> <OCCAM_LATITUDE_T_LONGITUDE_T_DEPTH_grid>
>> <link match="/lasdata/axes/d4733_restart1_a_nc_LATITUDE_T"/>
>> <link match="/lasdata/axes/d4733_restart1_a_nc_LONGITUDE_T"/>
>> <link match="/lasdata/axes/d4733_restart1_a_nc_DEPTH"/>
>> </OCCAM_LATITUDE_T_LONGITUDE_T_DEPTH_grid>
>> ...
>> </grids>
>>
>> <axes>
>> <d4733_restart1_a_nc_LATITUDE_T units="degrees_north" type="y">
>> <arange start="-78.125" step="0.25" size="577"/>
>> </d4733_restart1_a_nc_LATITUDE_T>
>> <d4733_restart1_a_nc_LONGITUDE_T units="degrees_east" type="x">
>> <arange start="0.125" step="0.25" size="1440"/>
>> </d4733_restart1_a_nc_LONGITUDE_T>
>> <d4733_restart1_a_nc_DEPTH units="layer" type="z">
>> <arange start="1" step="1" size="36"/>
>> </d4733_restart1_a_nc_DEPTH>
>> </axes>
>>
>> However, when I now try to plot this in LAS, the default full-region
>> x-y
>> plot for salinity, I get the error:
>>
>> Illegal region
>>
>> You specified a time or depth value that is outside the region
>> contained by this
>> dataset.
>>
>> Checking the LAS error log, I find the command it tried to run in
>> ferret was:
>>
>> set region/x="180":"540"/y="-78.0":"65.875"/z="1":"1"/l=1
>>
>> And the ferret error message:
>> **ERROR: illegal limits: "SALINITY" is not in the range X=180:540
>> Axis extremes are X=0.5:36.5
>>
>> Can anyone suggest a way to fix this so I can access this under LAS?
>>
>> Thanks very much,
>> - Andrew
>>
>>
>>
>> -----------------------------------------------------
>> Andrew Woolf (awo@mail.nerc-essc.ac.uk)
>> Environmental Systems Science Centre (ESSC)
>> Reading University
>> 3 Earley Gate
>> Whiteknights
>> Reading RG6 6AL
>> UNITED KINGDOM
>> Phone: +44 (0)118 931 8741 Fax: +44 (0)118 931 6413
>> -----------------------------------------------------
>
> --
> Joe Sirott
>
>
>
> ----------------------------------------------------------------
> \cancel mode verify
> ! init_modulo_x.jnl
> ! 9/00
> !
> ! The optional second variable is included for vector plots.
>
> ! Description: open (and initialize) a dataset and set the modulo
> ! flag on the time axes of the incoming variable(s).
>
> ! arguments: 1 2 3 4
> ! Usage: GO init_modulo_x [dset] [var num] [dset num] [var1]
> !
> ! dset -- the name of the dataset
> ! var num [1/2] -- first var or second (comparison) var
> ! dset num [1/2] -- first dataset opened or second
> ! (comparison var may be from first dataset)
> ! var 1 -- name of var 1
> ! var 2 -- name of var 2 (for vectors)
>
> ! e.g. yes? go init_modulo_x coads_climatology 1 uwnd
>
> ! check the arguments
> ! (var2 is optional)
>
> ! We cannot check argument 1 because absolute pathnames
> ! are parsed by ferret as command qualifiers. Checking
> ! for arguments 2 - 4 should be sufficient.
>
> set mem/size=20
> query/ignore $2%<Usage: GO init_modulo_x data_set var_# data_set_#
> variable_1%
> query/ignore $3%<Usage: GO init_modulo_x data_set var_# data_set_#
> variable_1%
> query/ignore $4%<Usage: GO init_modulo_x data_set var_# data_set_#
> variable_1%
>
> ! Hack to get a quoted string for DODS datasets.
> def symbol dset = "$1"
> set data ($dset)
>
> ! Set the x axis "modulo" flag
> define axis/x=0:359.75:0.25/units=degrees_east/modulo
> `$4,return=xaxis`
>
> ! We must define var$2_grid use in differencing
> ! This grid must reference a file variable: one that actually
> ! exists in dataset 'dset' as opposed to a virtual variable
> ! created with an expression involving other variables.
> define grid/like=$4[d=$3] var$2_grid_
>
> ! Any extra work: defining virtual variables, regridding, etc
> ! should be done at this time.
>
>
>
--
| NOAA/PMEL | ph. (206) 526-6080
Steve Hankin | 7600 Sand Point Way NE | FAX (206) 526-6744
| Seattle, WA 98115-0070 | hankin@pmel.noaa.gov
[Thread Prev][Thread Next][Index]
Dept of Commerce /
NOAA /
OAR /
PMEL /
TMAP
Contact Us | Privacy Policy | Disclaimer | Accessibility Statement