BreakPoint

cancel
Showing results for 
Search instead for 
Did you mean: 

Interesting Stats

I decided that I needed a little practice with regular expressions so I challenged myself to compose a single regex that would parse the useful data from the source of the forum pages.  It wasn't quite 'War and Peace', but there were a couple of lengthy expressions.  The side effect was that I generated a large dataset, and can now look at some statistics I couldn't before.

 

Kudos for 2009.

 

The distribution of Kudos for posts is very close to a Poisson distribution with an expectation value of 0.154 Kudos per post (I had to exclude the exploits of one individual whose weekly posts skew the distribution of high-Kudoed posts).  Some percentages:

 

Kudos (percentage)

0 (84.5)

1 (12.7)

2 (2.1)

3 (0.5)

4 (0.1)

>4 (0.1)

 

Once you hit 3 Kudos, you are in rarified air indeed.  Even 2 gets you an 'A+'

 

Solutions:

 

In 2009 there were roughly 13500 threads started, and 3911 solved (in 2009), a roughly 30% success rate.  Can you guess who the top 5 solvers were over the past 6 months (Aug 1 - Feb 1)?  The top 3 are no surprise, number 4 surprised me somewhat, and there is a three-way tie for fifth with two easy guesses and one interloper.

 

Interesting reads.

 

Eventually, I want to find ways to quantify interesting threads, a couple that I found during this exercise (hopefully an interesting stroll down memory lane).

 

Top Kudoed solutions:

http://forums.ni.com/ni/board/message?board.id=170&thread.id=411279

http://forums.ni.com/ni/board/message?board.id=170&thread.id=392472  (timeless classic)

 

Not-so-Kudoed

http://forums.ni.com/ni/board/message?board.id=170&thread.id=439600 (solved in first reply, followed by six pages of more replies)

 

Patience is a virtue:

http://forums.ni.com/ni/board/message?board.id=170&thread.id=107332 (only took 4 years to solve)

 

Finally Kudos/Post on LabVIEW board in 2009:

 

Top 5 for 2009 posts: (min 100 posts)

1. Darren 2.25

2. DFGray 1.03

3. Ben 0.76  (over 1800 posts!)

4. JarrodS  0.69 (sensing a trend)

5. Darin.K  0.66 (even a blind squirrel...)

 

followed by Andrey Dmitriev, altenbach (again sustained over 1700+ posts),mikeporter, Mathew Kelton,tst, and Ray.R

Message 1 of 19
(11,629 Views)

Darin.K wrote:
...(I had to exclude the exploits of one individual whose weekly posts skew the distribution of high-Kudoed posts)...

Hey, I represent that remark!

0 Kudos
Message 2 of 19
(11,625 Views)

Could you share some of the magic (regexes)?

I alway struggle my through them and alway have a feeling I can do it in a better way.

 

Ton

Free Code Capture Tool! Version 2.1.3 with comments, web-upload, back-save and snippets!
Nederlandse LabVIEW user groep www.lvug.nl
My LabVIEW Ideas

LabVIEW, programming like it should be!
0 Kudos
Message 3 of 19
(11,624 Views)

Darren wrote:

Darin.K wrote:
...(I had to exclude the exploits of one individual whose weekly posts skew the distribution of high-Kudoed posts)...

Hey, I represent that remark!


Maybe I was talking about dixy normous

 

Actually, 2+ Kudos/post is just plain sick.  well done.

 


TCPlomp wrote:

Could you share some of the magic (regexes)?

I alway struggle my through them and alway have a feeling I can do it in a better way.

 

Ton


I'll clean up some of my code, and post a VI with some sample text and the regex and see if we can't whittle away some more of it.  So much trial and error, mostly error.  The hard part is that the more complex you make it, the more suceptible you become to a single glitch throwing off the entire match.

 

0 Kudos
Message 4 of 19
(11,612 Views)

Darin.K wrote:
...Maybe I was talking about dixy normous

Any relation to Moxy Fruvous?

Message 5 of 19
(11,609 Views)

 


http://forums.ni.com/ni/board/message?board.id=170&thread.id=439600 (solved in first reply, followed by six pages of more replies)

 

 

Actually the point in the solution was explained in the replies and after a lot of going back and froth i made my point.......lol!!!

0 Kudos
Message 6 of 19
(11,566 Views)

Here is the VI I used to fine tune the regexes.  It will bold the text that matches in the control.  This text is extracted from a page listing threads, I basically clean the string (remove newlines) and find the individual threads by extracting the text between two tags (<tr> and </tr>).  The text here is for a solved thread with Kudos.

 

 

Message 7 of 19
(11,523 Views)

Thank you very much!

 

I set as  a goal of "0.5 over-all time" years ago to keep me from posting too much and keep the quality up. Once I reached it, I started using any Kudos over 0.5 as a license to post another reply. Since I only have access to the totals, I could not know what you have reported about the recent past so "a tip-of-the-hat" to you sir and again, thank you!

 

Is this something that can be run across all contributors and be posted as a spreadsheet?

 

Ben

Retired Senior Automation Systems Architect with Data Science Automation LabVIEW Champion Knight of NI and Prepper LinkedIn Profile YouTube Channel
0 Kudos
Message 8 of 19
(11,513 Views)

Something like this, all contributors who received at least one Kudo in 2009 (at least for a post in 2009).

 

 

Message 9 of 19
(11,505 Views)

Nobody wants to play the guessing game for the last six month solvers so I'll simplify a little.

 

1.  smercurio_fc (117)

2. Dennis Knutson (112)

3. Ravens Fan (102)

4. ??? (59)

5t. altenbach, Ben, Darin.K (57)

 

Who is the mysterious number 4?  It made sense once I knew, but it would have taken me a very long time to guess it.  Only mystery in the top 3 was the final order, and I was a little surprised to be interloping at 5, but it looks like ok company there.

Message 10 of 19
(11,481 Views)