NOTICE: This version of the NSF Unidata web site (archive.unidata.ucar.edu) is no longer being updated.
Current content can be found at unidata.ucar.edu.
To learn about what's going on, see About the Archive Site.
>From: "Todd Plessel" <address@hidden> >Organization: EPA >Keywords: 200311142115.hAELFUOb027573 netCDF reading Hi Todd- >Hope you're not buried in snow. Haven't really gotten any except in the mountains. Just lots of wind. >We have some large M3IO files - e.g., 80+ variables and hundreds of >time-steps. We notice that IDV takes a while to display the first >time-step of a single selected variable as a Color Shaded Plan View. >Examining the code it appears that all variables and all time-steps >may be getting read into memory. Is this true? >If so, it does not scale well with our data. If you are are talking about grids, then when you first open the file, it should just read in all the metadata. When you select a variable and display, all selected times are read in. You can subselect the times by: 1. Right click on the dataset name in the Data Sources panel and select the Select Times... menu. Use this to set the subset of times for all displays in this dataset. 2. In the Times panel, you can override the times set in 1 above. By default, whatever is set for the dataset is used. >It would be much better if only one time-step of the selected variable >is read at a time. Well, essentially it is, but you've got to tell it what times you want if you don't want all of them. If I'm missing something here, let me know. Don Murray