07-29-2010 04:43 PM - edited 07-29-2010 04:46 PM
I have a bookmark to http://culverson.com/keeping-your-charts-up-to-date/ this article explaines how to put a time stamp on the X axis of charts.
My company has installed a Barracuda web filter system and Barracuda blocks access to Culverson.com saying it has somthing to do with proxies. I am not sure what that means, but I do know I will get in big trouble trying to access this now since it is on the Barracuda black list.
Does anyone have a link to a simmilar explanation of how to put a timestamp on the X axis of a chart?
07-29-2010 06:51 PM
Relying on charts to plot a timestamp is inherently flawed. In my opinion, charts altogether are inherently flawed, except for "quick and dirty" troubleshooting, or where the time component of the measured signal is irrelevant (uncommon). With Charts, each subsequent point is assumed equally spaced, which may be a decent assumption for short periods of time with a deterministic data acquisition.
Instead, use XY Graphs, which allow for arbitrary point spacing in both X and Y directions. This means that over time, error in time measurement does not stack up, but instead always properly represents the absolute time and relative spacing between points.
To show the X axis as a timestamp, right click the XY graph to access the properties of that graph. Go to a tab that says something like "Scales" (I don't have LabVIEW in front of me) and then choose the X-axis scale from the Listbox. Here, you can set the Display Format of the numeric to be shown as an Absolute Time, and you can customize how this time is displayed.
I'm a bit surprised anyone would condone using a Timestamp on a Graph, and I would not recommend this practice when timing is relevant.
07-30-2010 09:25 AM
I will give the X-Y idea a shot today.
I am willing to say that 99.99999999999% of the data collection I do (and I do a lot) is done at regular intervals. In this case temperature data is being taken every fifteen minutes to determine temperature stabilization per UL1778. (No more than a 2 degree rise over three consecutive 15 minute intervals)
I really hate working with charts of any type in Labview and only use them as a rough visual indicator.
All "real" data analysis is done in Excel...