cancel
Turn on suggestions
Auto-suggest helps you quickly narrow down your search results by suggesting possible matches as you type.
Showing results for 
Search instead for 
Did you mean: 
ndvorak_forrester_com
‎09-01-2015
  • 9 Posts
  • 0 Likes Given
  • 0 Solutions
  • 0 Likes Received

Re: How to Transform Data Set from Long to Wide Across Multiple Variab...

by Calcite | Level 5 in SAS Programming
‎03-07-2014 10:44 AM
‎03-07-2014 10:44 AM
So here is what I came up with, but because a row is identified by the combination of the UserID and AppName, it does not work. I also think there must be a better way to transpose this data. PROC SORT DATA=WORK.wave1appdow;   BY AppName UserID; RUN; DATA transtest3;   SET wave1appdow;   BY AppName UserID;   KEEP AppName UserID        Mean_1-Mean_7        Sum_1-Sum_7        Freq_1-Freq_7       RETAIN Mean_1-Mean_7        Sum_1-Sum_7        Freq_1-Freq_7   ARRAY AppTimeMean (1:7) Mean_1-Mean_7;   ARRAY AppTimeSum (1:7) Sum_1-Sum_7;   ARRAY AppTimeFreq (1:7) N_1-N_7;   IF first.Appname THEN        DO;             DO i = 1 - 7;             AppTimeMean ( i )  = 0;             AppTimeFreq  ( i )  = 0;             AppTimeMed  ( i )  = 0;             END;        END;        AppTimeMean (D_DayofWeek) = Mean;        AppTimeSum (D_DayofWeek) = Sum;        AppTimeFreq (D_DayofWeek) = Freq;   IF last.Appname THEN OUTPUT;   RUN; Any Suggestions? ... View more

Re: How to Transform Data Set from Long to Wide Across Multiple Variab...

by Calcite | Level 5 in SAS Programming
‎03-06-2014 05:30 PM
‎03-06-2014 05:30 PM
Here is another segment of the dataset where you will see Websites repeating for certain days of the week. A row is identified by a unique DayofWeek, User, and Website combination: DayofWeek User Website Mean Sum Freq Saturday 243736 Bonus Box Tops 15 15 1 Saturday 243736 Flip Chip 601.55 6617 11 Saturday 243736 MyChart 100 100 1 Saturday 243736 PCH Sweeps 35 35 1 Sunday 243736 Flip Chip 158.6 793 5 Sunday 243736 PCH Sweeps 107.33 322 3 Monday 243736 Bonus Box Tops 10 20 2 Monday 243736 Cash Slots 31.5 441 14 Monday 243736 Flip Chip 273 2730 10 Monday 243736 Graphing Calculator 669.33 2008 3 Monday 243736 Hot Light 93.5 187 2 Monday 243736 PokerHouse 507.5 1015 2 Monday 243736 Tri-Peaks Solitare 493.5 987 2 Tuesday 243736 Battle Nations 820.67 2462 3 Tuesday 243736 Hot Light 15 15 1 Tuesday 243736 PCH Sweeps 101 303 3 Tuesday 243736 Slots Universe 741.67 2225 3 Tuesday 243736 Treasure Diving 510.83 14814 29 Wednesday 243736 Bonus Box Tops 30 30 1 Wednesday 243736 Cash Slots 91.286 1917 21 Wednesday 243736 Castle Clash 600.69 43250 72 Wednesday 243736 Lucktastic 809 809 1 Wednesday 243736 PCH Sweeps 25 50 2 Wednesday 243736 Slots Universe 25 50 2 Wednesday 243736 Treasure Diving 573.51 30396 53 Thursday 243736 Bonus Box Tops 5 5 1 Thursday 243736 Flip Chip 311.75 3741 12 Thursday 243736 Hot Light 640.5 1281 2 Thursday 243736 Lucktastic 121 121 1 Thursday 243736 MyChart 237 237 1 Thursday 243736 PCH Sweeps 155.5 622 4 Thursday 243736 Treasure Diving 738.35 22889 31 Thursday 243736 Tri-Peaks Solitare 238.25 953 4 Friday 243736 Flip Chip 671.67 4030 6 Friday 243736 Hot Light 982 982 1 Friday 243736 PCH Sweeps 181 181 1 Friday 243736 Tri-Peaks Solitare 164.5 329 2 Sunday 398526 Bloomberg 23 23 1 Monday 398526 Blackhawks 35 35 1 Saturday 398942 Spider Free 2 2 1 Friday 398942 Spider Free 30 30 1 Sunday 429762 DownloaderFREE 189 189 1 Sunday 429762 autoshutter 127 127 1 Monday 429762 Angry Birds Star Wa 290 290 1 Monday 429762 Toysrus 1 1 1 Tuesday 429762 Angry Birds Star Wa 1588 1588 1 ... View more

Re: How to Transform Data Set from Long to Wide Across Multiple Variab...

by Calcite | Level 5 in SAS Programming
‎03-06-2014 04:51 PM
‎03-06-2014 04:51 PM
I will need to summarize for each day of the week by company but for this example you are correct, only one day of the week for each company ... View more

Re: How to Transform Data Set from Long to Wide Across Multiple Variab...

by Calcite | Level 5 in SAS Programming
‎03-06-2014 04:36 PM
‎03-06-2014 04:36 PM
Yes that date is attached to a date, but the data has been summarized for each day of the week by Mean, Sum, and Freq. The numbers representing the Mean, Sum, and Freq variables is Time Spent on the website. I should have mentioned that the DayofWeek variable is numeric, running from 1=Sunday to 7=Saturday, and is formatted to show the day of the week using the FORMAT downame. ... View more

How to Transform Data Set from Long to Wide Across Multiple Variables

by Calcite | Level 5 in SAS Programming
‎03-06-2014 04:30 PM
‎03-06-2014 04:30 PM
Hello! Please help me transpose this data set from Long to Wide. Here is what I have: DayofWeek User Website Mean Sum Freq Saturday 17306100 MiKandi 10 10 1 Friday 17306100 Twitter 75 75 1 Sunday 17306100 Voice Search 151.67 455 3 Saturday 17306100 icomfort 30 30 1 Saturday 17306243 AOL 87 261 3 Sunday 17306243 Email 120.65 7963 66 Monday 17306243 Firefox 536 536 1 Tuesday 17306243 Google 15 15 1 Wednesday 17306243 IMDb 202.5 405 2 Thursday 17306243 Landry's, Inc 25 25 1 Thursday 18114509 Norton Mobile 45 45 1 Thursday 18114509 Notes 317.5 3175 10 Thursday 18114509 Our Groceries 510.94 8175 16 Friday 18072613 Chase 245 2450 10 Friday 18072613 Check 374.38 2995 8 Friday 18072613 Craigslist 55 55 1 Wednesday 18062931 Barcode Scanner 5 5 1 Wednesday 18062931 BofA 69.923 909 13 Wednesday 18062931 Facebook 343.94 35426 103 Here is what I need: Essentially multiple each day of the week (7 days of the week) by each of the 3 variables (Mean, Sum, Freq) for a total of 21 new variables. I want to keep the User and Website columns constant and unchanged. Could someone please help me attack this with a do loop and array, so that I am able to replicate it using the same procedure but for a WeekoftheYear variable (which holds 52 different values, one for each week of the year). Thank you! User Website MeanSunday MeanMonday MeanTuesday MeanWednesday MeanThursday MeanFriday MeanSaturday FreqSunday FreqMonday FreqTuesday FreqWednesday FreqThursday FreqFriday FreqSaturday FreqSunday SumMonday SumTuesday SumWednesday SumThursday SumFriday SumSaturday 17306100 icomfort 17306100 MiKandi 17306100 Twitter 17306100 Voice Search 17306243 AOL 17306243 Email 17306243 Firefox 17306243 Google 17306243 IMDb 17306243 Landry's, Inc 18062931 Barcode Scanner 18062931 BofA 18062931 Facebook 18072613 Chase 18072613 Check 18072613 Craigslist 18114509 Norton Mobile 18114509 Notes 18114509 Our Groceries ... View more

Re: Create Customized Time Intervals

by Calcite | Level 5 in SAS Procedures
‎02-20-2014 12:54 PM
‎02-20-2014 12:54 PM
Thank you very much! ... View more

Re: Create Customized Time Intervals

by Calcite | Level 5 in SAS Procedures
‎02-20-2014 10:31 AM
‎02-20-2014 10:31 AM
Hi Reeza, Sorry, I meant 3 additional columns. The columns that I want to create is the "Variable" column in the first table, and then 2 additional columns with "Day of the Week" and "Month of the Year". I included the first table to show the customized time interval by which I need to classify the time stamp. ... View more

Create Customized Time Intervals

by Calcite | Level 5 in SAS Procedures
‎02-20-2014 09:24 AM
‎02-20-2014 09:24 AM
Hello, I need to create customized time and date intervals for my time series data. I have a time stamp, and I want to classify the hours of the day into categories, as well as the day of the week. I need two additional columns, one that classifies the time into: Time Category Variable Time Interval Early Morning EM 5AM - 8AM Morning MO 8AM - 11AM Lunch LU 11AM - 2PM Afternoon AF 2PM- 5PM Evening EV 5PM - 8PM Night NI 8PM - 10PM Late Night LN 10PM - 2AM Graveyard GY 2AM - 5AM I also want to do the same with Days of the week (Sun-Sat) and Month (Jan - Dec). The time variable is LocalDateTime and is under the format DATETUNE. and looks like 03NOV13:14:42:33. My data looks like this: The data is called JanFile USER DEVICE LOCALDATETIME SECONDS_ACCESSED LOCATION MARKET 2133310 18116198 30NOV13:16:03:21 45 Away UK 2133310 18119318 03NOV13:12:46:56 95 Away UK 2133310 18119318 03NOV13:13:57:43 264 Away UK 2133310 18119318 09NOV13:08:25:12 1630 Away UK 2133310 18119318 09NOV13:09:53:04 1222 Away UK 2133310 18119318 09NOV13:09:23:31 1707 Away UK 2133310 18119318 09NOV13:08:07:03 1089 Away UK 2133310 18119318 03NOV13:12:45:10 75 Away UK 2133310 18119318 17NOV13:14:45:38 5 Away UK 2133310 18119318 23NOV13:22:58:06 917 Away UK 2133310 18119318 29NOV13:22:29:39 5 Away UK 2133310 18119318 04NOV13:08:02:17 119 Away UK 2133310 18119318 10NOV13:09:59:20 5 Away UK 2133310 18119318 15NOV13:16:57:02 5 Away UK 2133310 18119318 19NOV13:07:29:18 1722 Away UK 2133310 18119318 12NOV13:07:27:29 1277 Away UK 2133310 18119318 29NOV13:18:45:04 5 Away UK 2133310 18119318 25NOV13:07:28:52 1734 Away UK 2133310 18119318 27NOV13:08:00:33 15 Away UK Please help!! Thank you! ... View more

Creating an Identifier Variable

by Calcite | Level 5 in SAS Procedures
‎02-19-2014 03:01 PM
‎02-19-2014 03:01 PM
Hello, I need help creating an Identifier Variable where the value is either 1 or 0 depending on a certain condition. Here is my sample dataset: USER ID   DEVICE ID 1               122     1               239 1               239     1               239     2               144 2               144     2               144     2               223 2               223 3               111 3               111 3               111 4               155 4               155 I need to add an additional variable called "BOTHDEVICES" if a user has 2 or more devices = 1 or if there is only data for 1 device = 0. Each device has a unique ID number. In this example, Users 1 and 2 would have BOTHDEVICES values of 1 and users 3 and 4 would have BOTHDEVICES values of 0, which would be repeating as shown below: USER ID   DEVICE ID      BOTHDEVICES 1               122                    1 1               239                    1 1               239                    1     1               239                    1     2               144                    1 2               144                    1     2               144                    1     2               223                    1 2               223                    1 3               111                    0 3               111                    0 3               111                    0 4               155                    0 4               155                    0 Thanks for your help! , N ... View more
Contact Me
Online Status
Offline
Date Last Visited
‎09-01-2015 07:11 AM
Quick links