Date: Sun, 7 Jan 2007 11:30:37 -0500
Reply-To: Arthur Tabachneck <art297@NETSCAPE.NET>
Sender: "SAS(r) Discussion" <SAS-L@LISTSERV.UGA.EDU>
From: Arthur Tabachneck <art297@NETSCAPE.NET>
Subject: Re: Unofficial 2006 Year-End Review - CORRECTION
Content-Type: text/plain; charset=ISO-8859-1
There was an error in the original post which I sent in response to Hari's
question. It didn't seem reasonable that so many posts were listed as
being on Sunday and, of course, there was an error in the code.
Specifically, day name was based on a day variable that had been formatted
as being weekday rather than date. While SAS accepted the code without
error or warning, all days were one off. Thus the correct response to
Hari should have been:
Hari,
While I had indicated that the list's unrepresentative date-time stamps
precluded their use for the kind of analysis you suggested, I did observe
that both Toby's and David's posts had consistent date-time stamps (albeit
Toby's had to be adjusted because he really isn't from England), and
wondered if an analysis of their posts might answer your question.
Result: we definitely can't use David's posts, at least as far as times
are concerned. Interestingly, he either really does sleep, or at least
elects not to respond to posts between 1 and 7 am, and seldom before 10
am! However, most of David's posts are between 10 pm and midnight, which
I don't think reflects the majority of the list's activities.
Both Toby's and David's posting activity results do concur with WEDNESDAY
being the day when most people post, and SATURDAY being the day that the
fewest posts are recorded, followed by SUNDAY.
Toby may well be a 'bot', as David has suggested sometime ago, as the only
hours he doesn't post are between 4 am and 6 am. However, it is also
possible that he is doing something other than sleeping then, there is
also the possibility that there simply isn't anything requiring a response
at those times. If Toby really does reflect a microcosm of the list's
posts, then the majority of posts occur between 8 am and 11 am, and then
taper off during the day and are virtually non-existent during the time
when most of us sleep.
In an off-line conversation (I did get David's and Toby's permission to
post this level of detail), David suggested that his daily counts might
not reflect the entire lists' behavior. As such, I did a day-of-week
analysis for all of the list's 2006 posts. As you can see, SATURDAY really
does appear to be the day with the fewest posts, followed by SUNDAY.
Activity picks up again on MONDAY and WEDNESDAY is THE day with the most
posts.
If this still doesn't answer your question, let me know what might.
Art
Results
-------
Name Day Count Percent Hour Count Percent
---- --- ----- ------- ---- ----- -------
David
Sun 278 11.01 0 62 2.45
Mon 363 14.37 1 5 0.20
Tue 459 18.17 8 3 0.12
Wed 551 21.81 9 1 0.04
Thu 451 17.85 10 52 2.06
Fri 351 13.90 11 209 8.27
Sat 73 2.89 12 171 6.77
13 222 8.79
14 210 8.31
15 161 6.37
16 127 5.03
17 113 4.47
18 24 0.95
19 28 1.11
20 51 2.02
21 203 8.04
22 571 22.60
23 313 12.39
Toby
Sun 96 4.80 0 22 1.10
Mon 364 18.20 1 9 0.45
Tue 362 18.10 2 3 0.15
Wed 397 19.85 3 3 0.15
Thu 370 18.50 7 4 0.20
Fri 377 18.85 8 222 11.10
Sat 34 1.70 9 249 12.45
10 200 10.00
11 182 9.10
12 140 7.00
13 147 7.35
14 159 7.95
15 143 7.15
16 170 8.50
17 40 2.00
18 42 2.10
19 42 2.10
20 72 3.60
21 69 3.45
22 54 2.70
23 28 1.40
Summary for the entire list
Sun 1756 5.17
Mon 5252 15.46
Tue 6221 18.32
Wed 6784 19.97
Thu 6538 19.25
Fri 5978 17.60
Sat 1435 4.23
On Tue, 2 Jan 2007 05:27:27 -0800, Hari <excel_hari@YAHOO.COM> wrote:
>Art,
>
>Just for fun.. I was wondering as to whether the posts fluctuate a lot
>by day of the week (more on Monday/Tuesday than rest etc?). Also,
>whether it changes much by time of the day whether more during working
>hours than outside of it. But yes, it might be difficult to do this
>accurately as folks post from different time-zones or can that be taken
>care of?
>
>Regards,
>HP
>India
>
>Arthur Tabachneck wrote:
>> First of all, happy new year to all.
>>
>> I couldn't resist. During 2006 the list increased its number of posts
by
>> 26%, from 26,992 in 2005, to 33,963.
>>
>> And, not all due to David, as his posts only increased by 18.4%, from
>> 2,133 to 2,526. However, that was enough to keep him, again, as the
>> list's most frequent poster.
>>
>> Thus, until Howard gets a chance to verify the numbers, here are the
>> unofficial stats.
>>
>> 50 Most frequent posters (actually 51 as there was a 3-way tie for 49th
>> place):
>>
>> rank name COUNT
>> 1 David Cassell 2526
>> 2 Toby Dunn 2000
>> 3 Howard Schreier 1519
>> 4 Arthur Tabachneck 618
>> 5 Jim Groeneveld 604
>> 6 Gerhard Hellriegel 587
>> 7 datanull@gmail.com 558
>> 8 Joe Whitehurst 536
>> 9 Alan Churchill 490
>> 10 Richard DeVenezia 487
>> 11 Paul Choate 470
>> 12 Jiann-Shiun Huang 448
>> 13 Mark Terjeson 422
>> 14 Sig Hermansen 397
>> 15 Ya Huang 389
>> 16 Ron Fehd 378
>> 17 Peter Flom 371
>> 18 Ian Whitlock 326
>> 19 Peter Crawford 301
>> 20 Roland Berry 299
>> 21 Venky Chakravarthy 276
>> 22 Kevin Viel 264
>> 23 WenSui Liu 226
>> 24 Nathaniel Wooding 224
>> 25 Michael Raithel 221
>> 26 Sté°¨ane Colas 218
>> 27 Clinton Rickards 194
>> 28 Jack Hamilton 189
>> 29 Yu Zhang 188
>> 30 Roy Pardee 176
>> 31 Shiling Zhang 171
>> 32 Hari Prasadh 169
>> 33 Ben Powell 168
>> 34 Paul Dorfman 167
>> 35 proccontents@gmail.com 153
>> 36 Harry Droogendyk 152
>> 37 Eric Hoogenboom 146
>> 38 Dale McLerran 144
>> 38 Mehdi Soleymani 144
>> 40 Rune Runnestoe 136
>> 41 Ken Borowiak 134
>> 42 Daniel Nordlund 132
>> 43 Ken Lin 131
>> 44 John Gerstle 116
>> 44 Monal Kohli 116
>> 46 Hari Nath 113
>> 47 Madan Kundu 112
>> 48 Chang Chung 99
>> 49 Irin Figvam 97
>> 49 Muthia Kachirayan 97
>> 49 Robert Bardos 97
>>
>> While number of lines (as calculated by the listserv) may not be the
most
>> accurate definition of post length, using it as a measure the following
>> either had a lot to say or .. The envelope please (note: only
considered
>> those who had at least 50 posts):
>>
>> rank name Average # of lines
>> 1 Randy Herbison 167
>> 2 Bora Yavuz 161
>> 3 Joe Whitehurst 143
>> 4 Jake Bee 122
>> 5 Mark Terjeson 122
>> 6 Jiann-Shiun Huang 120
>> 7 Paul Dorfman 119
>> 8 Nathaniel Wooding 118
>> 9 Jack Clark 115
>> 10 Irin Figvam 114
>>
>> So, who was either extremely succint, didn't paste the entire thread,
or ..
>> The following are the 10 (who had at least 50 posts) who had the lowest
>> average number of lines:
>>
>> rank name Average # of lines
>> 1 Tanwan Zang 27
>> 2 Dianne Rhodes 31
>> 3 Mehdi Soleymani 32
>> 4 Eric Hoogenboom 33
>> 5 slhappyls@gmail.com 36
>> 6 Dimitri Shvorob 41
>> 7 Roland Berry 42
>> 8 A.Benedictus 42
>> 9 Byron Kirby 43
>> 10 Mkrogh@dsr.kvl.dk 45
>>
>> Which words showed up most frequently in the subject line? Excuding
>> commonly used words like a, the, I, etc., the 20 words which showed up
>> most frequently were:
>>
>> rank word count
>> 1 HOW 2447
>> 2 MACRO 2373
>> 3 QUESTION 1690
>> 4 VARIABLE 1621
>> 5 HELP 1284
>> 6 SQL 1204
>> 7 PROBLEM 1040
>> 8 VARIABLES 1025
>> 9 FILE 1012
>> 10 USING 979
>> 11 DATASET 888
>> 12 VALUES 713
>> 13 VALUE 640
>> 14 DATE 627
>> 15 OUTPUT 614
>> 16 EXCEL 593
>> 17 FORMAT 566
>> 18 INTO 544
>> 19 ABOUT 537
>> 20 BY 515
>>
>> When did we post the most? March!
>> When did we post the least? December!
>>
>> month count
>> March 3182
>> May 3136
>> October 2979
>> August 2909
>> July 2893
>> January 2842
>> November 2837
>> April 2811
>> June 2745
>> September 2720
>> February 2526
>> December 2383
>>
>> Further, we seem to be middle-of-the-week people:
>>
>> week count
>> 3 8199
>> 2 7931
>> 1 7716
>> 4 7547
>> 5 2570
>>
>> Again, these results are unofficial but, in the chance you were curious
>> too, I thought I'd share them.
>>
>> Art
| 
