This archive contains answers to questions sent to Unidata support through mid-2025. Note that the archive is no longer being updated. We provide the archive for reference; many of the answers presented here remain technically correct, even if somewhat outdated. For the most up-to-date information on the use of NSF Unidata software and data services, please consult the Software Documentation first.
Hi Pete, > metadata=> select * from grid_info limit 10; > id | datasetid | ensembleid | secondaryid | level_id | location_id | > parameter_abbreviation > ----+------------------+------------+-------------+----------+-------------+------------------------ > 1 | FE_GFS_Globalp25 | | | 44 | 30 | Vis > 2 | FE_GFS_Globalp25 | | | 407 | 30 | uW > 3 | FE_GFS_Globalp25 | | | 474 | 30 | REFD > 4 | FE_GFS_Globalp25 | | | 1 | 30 | REFC > 5 | FE_GFS_Globalp25 | | | 45 | 30 | PMSL > 6 | FE_GFS_Globalp25 | | | 407 | 30 | vW > 7 | FE_GFS_Globalp25 | | | 551 | 30 | PVV > 8 | FE_GFS_Globalp25 | | | 407 | 30 | VRATE > 9 | FE_GFS_Globalp25 | | | 44 | 30 | WGS > 10 | FE_GFS_Globalp25 | | | 474 | 30 | SNMR > (10 rows) So, this above output is showing the column names in order (id, datasetid, ensembledid, etc) and then 10 records with their respective information (some of the columns may be null, as shown in this output). > metadata=> delete from grid_info where id=FE_GFS_Globalp25; The command you tried to run was saying delete the record where the id equals "FE_GFS_Globalp25", but the id column consists solely of numbers (listed 1-10 in the output you showed above). I want to clarify quickly, this is not all the data that's in your table. Are you trying to remove quarter degree GFS data? Because that's what your delete command was attempting to do. There are likely thousands of rows in this table. You can see exactly how many rows with this command: select count(*) from grid_info; If you spend a little time googling around "psql select statements" you might be able to better narrow down exactly the records you're looking for. You only got 10 results because the command I first gave you limited it to 10 results ('limit 10'). One that might be useful is to grab all the unique values of the datasetid (in this case the first 10 entries are all the GFS Global .25 data. I think this should give you all the unique values of datasetid: select distinct datasetid from grid_info; This should list all the datasetids and hopefully the one you want to remove (conusarw) shows up as one of those ids? Then you'd need to either actually use the integer values of the ids and delete based on id, or you could do a string matching based delete, which I believe would be delete from grid_info where datasetid ilike '[enter dataset id]'; *note there would be no square [] brackets, I'm just using that to note where to put the name So if you did want to delete all GFS quarter degree you'd do delete from grid_info where datasetid ilike 'FE_GFS_Globalp25'; Granted, doing it this way doesn't allow you to keep track of the info table's id value, which would be used to delete the corresponding records from the grid table. But honestly, like I said previously, I think that data will be purged anyway and probably isn't a huge waste of resources. Let me know if this is making more sense? --Shay Carter She/Her/Hers AWIPS Software Engineer UCAR - Unidata If you're interested, please feel free to fill out a survey about the support you receive: https://docs.google.com/forms/d/e/1FAIpQLSeDIkdk8qUMgq8ZdM4jhP-ubJPUOr-mJMQgxInwoAWoV5QcOw/viewform Ticket Details =================== Ticket ID: YED-418649 Department: Support AWIPS Priority: Normal Status: Open =================== NOTE: All email exchanges with Unidata User Support are recorded in the Unidata inquiry tracking system and then made publicly available through the web. If you do not want to have your interactions made available in this way, you must let us know in each email you send to us.