1D ANOVA

[vehnasa]

 I am having a problem with implementing the 1D ANOVA. I am running a aircraft scenario and collecting data generated from this scenario. The data collected are the Aircraft ID and the that the aircraft crosses a certain waypoint. Now the scenario was run a total of 450 times. I have created an array that has the times across the waypoint, which is (X) in the ANOVA vi. I have determined that the number of levels is five. Am I going to have to match each one of the 450 values to a corresponding level, or is there a vi that will do that programatically? Am I even using the ANOVA properly? I have looked at all the examples on the ANOVAs and tried some variation of them but always, and this is a biggy, returns an "NaN" no matter what output I am looking at from the ANOVA. Can anyone help me with this??? I will post a code snippet as soon as find out how.

[StephieU]

I am currently teaching myself ANOVA through LabVIEW so if there is nobody more qualified than perhapse I can try to help.

Here is my prototype code in case it helps at all.  The data input is set up this way since it will run in parallel to other code to make sure the samples are indeed from the same population (no problems with machining specimens or doing data collecting on different days) before the data is combined into one set and the bulk analysis done.  If you need LabVIEW to select your indices then you can perhapse augment the rainfall example to your purposes.

I was taught statistics and ANOVA through Primer of Biostatistics by Stanton Glantz and have used the simple web-based program One-Way Analysis of Variance for Independent or Correlated Samples to make sure the ANOVA output is making sense.  All together I have found the NI support to be lacking in this area although it is usually stellar for my other coding purposes.

[StephieU]

Message Edited by StephieU on 05-07-2008 05:49 PM

[vehnasa]

Stephie,

thanks for the help. I would like to view your example but am using

LabVIEW 8.2.1 and your version is in 8.5. If you come up with some

other insight I would greatly appreciate it.

[Mathan]

Hai,

Here it is saved in ur version.

Thanks,

Mathan

[Karunya_R]

The 1-D Anova function is availabe in LabVIEW but just not in the base package. Here is the help file for this

[StephieU]

I think you need a new help file as everyone who as asked about ANOVA before in the discussion forum 1) already had the vi and 2) stated the help file provided no help.  Searcing other sites for help only provided journal articles where the data was first collected and then ANOVA was done with a second program.  Getting your ANOVA vi's more user-friendly (perhapse as an express vi or a more detailed ANOVA user manual) would make LabVIEW an even more powerful tool.

Stephanie Uhlich

[vehnasa]

I have the help file for the 1D ANOVA and have follwed its outline. Ican not for the life of me understand

why my results are returning NaN. To the NI application engineers that monitor this forum is there any one

that has put the ANOVAs to use and can provide more than the help to give direction on the application of

this VI. I have also studied the examples and have even spliced a bit of my program into it. Thanks everyone.

[StephieU]

Your 1D ANOVA.vi (as posted above) is not working because there needs to be more than one value for each index.  I played with some different scenarios and included is different things that must be done to get NaN.  This may be why it is not working but I am not sure of what your larger code looks like so you may have already solved this problem.

Hopefully the save for previous version worked.

[vehnasa]

Stephie,

I want to thank you and everyone-else. The last reply did the trick.

But can someone tell me why changing the values of the levels

effects the output of the ANOVA?? If I go from a 1 to a 3 on a level

value it could return an NaN. Thanks again to everyone who sent a

response to my disstress signal.