The LIST

[chilly_charly]

JLV
According to the latest List developments (to appear as soon as I solve a nasty cosmetic problem...), you are clearly among the top contributors of this forum. Quantity is not always a decisive criterium !
Don't focuse on the 1 K level. Look at JPD or RolfK... 😉
The rest is just a matter of time and disponibility. Nothing really to be proud of... except if you are able to keep it that way for years, instead of months !

CC

[Ray.R]

Well said CC!

It's not how many answers are produced but rather how helpful they are.

I must admit that although tst came out of nowhere, she's... (I meant "it") ... quite the contributor. Impressive.
And knowledgeable.. not to mention pretty. 😉

😄

[chilly_charly]

Want to see how enthusiastic you are ?

Just use the attached prototype vi (LV 7.0)

Far from finished, but since I'll be quite busy in the next few days, I prefer to release it as it is.

First, select the pathname for a recent Enthusiast file, then run.

You can either drag the cursor to identify a specific point. Not always easy, you may have to rotate the graph for that.
Or you can clik on the user name in the table, and the cursor will jump to the corresponding position in the graph.

I believe some of you will enjoy that !

This time, I have tried a more scientific approach. There are many variables which measure an Enthusiast's activity (some, such as the n° of given stars, or the number of connections, can't be accessed by a standard user).
Since the number of estimators is far over my own possibilities of multidimensionnal perception, I have used PCA (Principal Component Analysis). For those of you who never heared of this statistical tool, let me just say that it's a standard method to define the best combination of variables able to represent most of the variance of a complex set of data. You got it ?
No ? OK, to start with, you could have a look here

Waiting for your comments...

CC

[tst]

I'm missing some analysis VIs (plus, the graph became invisible again).

[chilly_charly]

Sorry tst !
I have checked (before posting) that the display and vis where OK on my computer. Don't know where the problem is...
If somebody else could test it, we could chase the BUG.

CC

[tst]

Try including the vi.lib VIs in the LLB.
Also, I was missing a couple from your original program. One I had (Excel get data...) and one I don't (Row Col to range format [excel vi?]).

[chilly_charly]

Here is a version with vi libs included. of course it's much bigger.

CC

[Ben]

Hi CC,

I had to select the "value2" in one of the Excel property nodes, but after that, it ran fine (LV 7.1).

Very interesting analysis. What are the units used for each axis?

Ben

[shoneill]

Hey CC,

I hope you're getting some comission from NI. I do believe I've found the very first compelling reason to upgrade beyond 6.1. Or should I still wait until 8.0 hits the shelves?

Shane.

PS Has anyone ever actually SEEN LV on "the shelves"? Does NI only sell direct?

[chilly_charly]

Ben,
the axis are adimentional : the PCA method begin with a normalisation step of all the adressed variables : division by the mean value, and correction by the standard deviation, just to have zero centered data, with a standard deviation equal to 1. Then it looks for the best "directions" in a n-dimentionnal space (eigen-vectors). The data are then rescaled in the new coordinate set. Basically, each axis is a linear combination of all the variables. I'll give you some more infos later... but I think that you can make some guess by locating a few members on the 3D representation.

I could be proud, since I'm apparently the most excentric point !... I think that everybody here agree with that statement ! 😄

Since the method can be applied to all the classification problems (as a very first step, since it's really basic and better tools do exist...) I hope to start a profitable discussion about multivariate analysis !

shoneill,
I'll try to post a LV 6.1 version later this evening. Just to allow you to wait for 8.0 😉

CC