Loading a whole bunch of files
_sk
I am trying to load a lot of files and obviously this takes a lot of memory. I tried sharing the files with the current experiment (as compared to copying) and I still stuck.
I can write a short script displaying the files one by one (as I have to sift through all this data one by one) and go around the issue.
Any other ideas as to what can be done here?
best,
_sk
Or load the files on-demand. Use the Open /MULT=1 to get the list of files you want to work on, but don't actually load the file until it is needed and kill it before the next file is loaded. It all depends on what you are doing with your data.
March 2, 2017 at 09:15 am - Permalink
If you're running out of memory, you could try using a 64-bit version of Igor. If you're still using IP6, and you're on Windows, you could try Igor Pro 6 64-bit (http://www.wavemetrics.net/Updaters/ReadMe%20(64-bit).html). If you're using Igor 7, it ships with 32-bit and 64-bit versions on both platforms. Just run the 64-bit version and that may solve your problem.
March 2, 2017 at 09:20 am - Permalink
If not, you need to think about the data types. Do you need floating point data (maybe real is enough) or can you use integer waves? Maybe you can even use integer waves with 16bit?
March 2, 2017 at 09:24 am - Permalink
The particular situation I was faced with was loading multiple small files beyond the amount of physical RAM.
Luckily I don't need all the files to be loaded at the same time, so I will just swap out to and from disk unneeded files and fit within RAM.
best,
_sk
March 3, 2017 at 05:02 am - Permalink