how can i view data files and print them?

[mattroos]

I'm collecting very large data files with the DAQ. Afterwards I'd like to be able to view them (zooming in and out and scrolling) then print when I get a view I like. How can I do this? Is there code out there already? What are some ways of easily visualizing large data files without using tons of PC memory and waiting for every data point to load?

[mikeporter]

There are some basic techniques that are useful in this sort of application. First, to get an overview of an entire dataset, a good place to start is to read in the file 20 or 30 samples at a time. As you read each block, perform an average and only plot the average. This on-the-fly decimation allows you to see very large blocks of data without bogging down you system.

Second, for the zoom function use the cursors. When the user selects an area to zoom in on, read just the values between the indices the user selects. If the selection is small enough you can display the data directly, otherwise you can decimate it as before-though probibly using a smaller block size.

For reading the position of the cursors, use property nodes.

The basic thing to rem
ember is that most PC screens are at most a couple thousand pixels wide. This video resolution places an absolute limit on the number of datapoints that you will be able to see without aliasing. In general it's better to decimate the data yourself than to let the video card do it.

Let me know if you have any specific questions

Mike...

[DennisM]

Mike

I am currently working on a project that requires display of datalog file on a graph. The datasets can hold up to 1 million samples. I have seen that displaying a complete dataset on a graph results in the CPU usage of the PC going through the roof, even though once the data is loaded the vi does nothing. I am keen to know more about how to handle the display of large datasets on a graph.

With the suggestions you mention here - when you zoom in on a dataset, presumably you 're-decimate'the data in order to display it. This implies that all of the data is read from the file up front. Is this correct? or do you also recommend that when reading the file a decimated set of samples is extracted.

Dennis