About jswislar

jswislar · ‎01-26-2018

Thank you for the quick response. Would the new cluster variable be entered as a "Nest" variable in GLM? I already know there are 4 clusters I can group the states into.

jswislar · ‎01-26-2018

I am looking to predict the proportion of patients in a hospital with a specific disease, adjusting for hospital characteristics. However, the data are clustered (hospitals clustered within states) and I want to adjust for this, too. I am using JMP Pro 13. Is the proper approach to use a Generalized Linear Model and select state with the "Nest" button? This seems correct, but when I do that JMP includes each state (n=51) as a variable. This is effectively what I want, but adds a lot of DF and variables to a data set with relatively few observations. Am I doing this right, or is there a better way? Thanks

jswislar · ‎10-26-2017

Thanks again, Jim. I've used many more rows than this. It was just a hypothesis. Every case of DX1 ends in a number, but the number of spaces between what I'm extracting for Var2 and that number varies. Maybe that was causing it. Either way, your solution worked.

jswislar · ‎10-26-2017

Thanks, Jim. That is how I tried it previouslly. For some reason it seems to crash JMP Pro 13 when I run it. Maybe because my data table is 65,000 lines?

jswislar · ‎10-25-2017

Thanks, Jim. That is what I was looking for. How do I run this script? I tried it as a column formula. That didn't work, but I didn't expect it to. I also opened a script window and ran it there, but it just stopped responding.

jswislar · ‎10-24-2017

The only rule that I can think of that would be consistent would be that the second word appears beween numbers (#). Ultimately I want the phrase between the hyphens. So, "#-word -#" and "#-word-word -#" and "#-word-word word -#"

jswislar · ‎10-24-2017

Is there a way to get WORD to ignore some dashes? Some of the second words I'm trying to extract are hyphenated, so I'm only getting a partial extraction. For example, for "938-ex-marine" I get "938" and "ex" rather than "ex-marine"

jswislar · ‎10-24-2017

Thanks! That works well.

jswislar · ‎10-24-2017

I want to extract character strings from a variable (Dx1), but the length and position of the strings varies, but they are separated by dashes. Sometimes there are multiple dashes. For example... "7812-abcdef -9" "v7892-xyz -0" "812-fghkj -0" "17361-gf-jhyt -9" What I would like to do is create a new variable that extracts the digits/characters before the first dash, and then another variable that extracts the characters between the first and second dashes. So, Var1 Var2 7812 abcdef v7892 xyz 812 fghkj 17361 gf-jhyt I tried following the example posted here, but it just extracted the dash. Changing the +1 to another number extacts that number of characters, but my data length varies. https://community.jmp.com/t5/Uncharted/JSL-Character-String-Functions/ba-p/21323 Here's what I wrote. What am I doing wrong? Trim( Substr( :Name( "Dx1" ), Contains( :Name( "Dx1" ), "-" ), 1 ) )

jswislar · ‎08-17-2017

Hmm. I ran it again and it works. Maybe it got hung up halfway through running the script. I can't figure out how to delete this post.

jswislar · ‎08-17-2017

This formula *seems* OK to me, but it's not giving me what I want, so something is wrong. I'm trying to say that if: Var1=. and Var2 >= 0 (i.e., not missing), then Var3=1 Var2=. and Var1 >= 0, then Var3=2 else (if Var 1 & Var2 are not missing) Var3=3 What I get is Var3=3 when Var 1 & Var2 are not missing as expected, BUT Var3=1 if EITHER Var1 OR Var2 are not missing. There is only one case where Var3=2 even though there should be a LOT more. That case is Var1=0, Var2=. Here is the formula for Var3: If( Is Missing( :Name( "Var1" ) & :Name( "Var2" ) >= 0 ), 1, If( Is Missing( :Name( "Var2" ) & :Name( "Var1" ) >= 0 ), 2, 3 ) )

jswislar · ‎08-10-2017

Thanks. The only way that I've been able to save my files is to find a new location for them each time I want to save them, then go back and duplicate the old file and move the new version back to the file it belongs in. I was hoping that there would be a more permanant solution.

jswislar · ‎08-09-2017

Does JMP create temp files while saving large files? I keep getting an I/O not enough disk space error message when I try to save a file. I am saving a very large file to an external server with lots of free space, so I know there's disk space. However, my copy of JMP resides on a hard drive with no space (maybe 0.5% available; C:). I'm thinking that this is the problem. My computer has another drive (E:) with gobs of space. If hard drive space is an issue, can I direct JMP to write its temp files to E:? IT set up the machine and installed the software, and I think getting them to move it will be a long laborious process. I already asked them for help and they just shrugged. Thanks!

jswislar · ‎07-28-2017

Thanks! I've been trying to learn JMP programming and this interaction helps, especially when odd requests aren't in the books.

jswislar · ‎07-28-2017

Great! Thank you for the clarification and the revised fomula! I appreciate it. The revision works EXCEPT if the missing pattern is mixed. For example, see "Site=001" below, my results look like this with the revised formula (Count) and what I want it to give me. Site date Var1 Count What_I_Want 001 May . . . 001 Jun 10 1 1 001 Jul . . . 001 Aug 12 . 2 001 Sep 15 . 3 002 May 13 1 1 002 Jun 14 2 2

jswislar · ‎07-27-2017

Jim, Thanks for following up. 1. Can we make theCount column stop counting when the value is missing? In you example, it repeats the last count number, will will presummably do so until the ID number changes. 2. I copied and pasted your formula. The only thing I changed in your formula is the name of the variables so that it exactly matches my data set. 3. The word "count" does not appear in any of my variable names. I renamed my "TheCount" variable to "Column 10" and got the same error message. I've also tried renaming my site variable. Same problem. 4. I'm using JMP Pro 13.1 Is Missing (Var1) == 0 Then TheCount=1 Translates as if Var1 is missing then the new variable (TheCount)=1? But what does "==0" do? This is what I have, including my variable names. If( Lag( :SiteID ) != :SiteID, If( Is Missing( :NozeroValue ) == 0, :Column 10 = 1, :Column 10 = . ), If( Is Missing( :NozeroValue ) == 0, :Column 10 = Sum( :Column 10, 1 ) ) ); :Column 10;

jswislar · ‎07-24-2017

Hmm. Same error message. I made up the data I gave you, but it's the same idea, except ID is character in my data. Would that make a difference? Also, noticed that your formula stopped counting at 2. Thanks!

jswislar · ‎07-24-2017

This formula gives me the following error "Illegal Reference (recursion) 1 times At rows:2" I follow the formula, but I can't figure out where the error is.

jswislar · ‎07-24-2017

I have a longitudinal study with sites entering on different dates. To make them comparable, I need to assign each entry date as "1" and the second data collection point as "2" and so on. The data are in long format and there are a lot of missing data (missed data collection) The data look like this, I want to generate "Count": ID Date Var1 Count 001 Jan-1 . . 001 Feb-1 . . 001 Mar-1 10 1 001 Apr-1 13 2 002 Jan-1 8 1 002 Feb-1 10 2 002 Mar-1 . . I used a JRP formula, but it counts every line whether the first measure is missing or not.

jswislar · ‎07-20-2017

This gives me the same thing as the first solution (# of potential data points), but in a very large data table. However, If I take your first solution and add the Data variable as the third grouping variable I still get an enormous table, but I do get a count of missing reports. I can work backwards from there. Thanks.

jswislar · ‎07-20-2017

Thank you, but this solution only gives me the number of potential data points, not the number of valid entries. For example, there were 15 opportunities for Site C to submit data for Report X (15 rows per site). Site C submitted 3 of 15 reports. The Summary Table solution you offered gave me Site C=15. It should be Site C=3.

jswislar · ‎07-20-2017

I have a dataset where sites reported data monthly, or they were supposed to. So, some sites didn't submit data for some months. I'd like to count up the number of valid reports by site to see if compliant sites differed from non-compliant sites. The data are in long format, such that... SITE REPORT MONTH DATA A ABC June 10 A ABC JULY . A ABC Aug 3 A XYZ June . A XYZ June 15 B ABC June 6 B ABC June 2 etc I'd like to end up with a data so I can say that Site A 2 valid ABC measures, and 1 valid XYZ measure, and then take those numbers and compare them to Site B. Thanks

jswislar · ‎06-28-2017

There's a "combined" group in my survival output. This is the survival time with all groups combined. OK, but how is it that the survival time for this group exceeds the survival for any of my individual groups? Shouldn't it be more of an average? Here are my results. Group # Failed # Censored Mean 1 182 1981 1360.99 2 820 9736 1285.93 3 254 4072 1263.10 4 49 616 1168.80 Cmbnd 1305 16405 1369.43

jswislar · ‎06-01-2017

I have a data set with multiple lines of data per individual. The lines of data are medical procedures and I want to get the average number of lines (or procedures) per individual. I'm not sure how to do that. I was able to count the number of lines but I can't take the mean of that since everyone who has 5 lines also has 4 lines and 3 lines and so on. So basically what I think I need is for JMP to count the number of lines, but then only save the highest number. How can I do that? The data structure is like this: ID Count 001 1 001 2 001 3 002 1 003 1 003 2

jswislar · ‎06-01-2017

Is it possible to account for clustering in survival analysis (Marginal hazards) in JMP Pro? I'm currently modeling procedure failure rates per tooth, but teeth are clustered within individuals, so I think the analyses should account for this. Can this be done with JMP? I don't see any options in the proportional hazards selections, but there is a "grouping" option in the simple survival analysis. Thanks

jswislar · ‎03-17-2017

Ahhh, you're right! I was thinking that only the lower order terms needs to be included. Thanks!

jswislar · ‎03-17-2017

I'm trying to create a model with a 3-way interaction term. I selected the variables and clicked on "cross" per JMP instructions for 2 or more variables. But JMP gives me the following warning: "The model is missing an effect X1*X2*X2*X1" Which is especially odd since what I entered into the model, and what appears in the dialog box, is: X1 X2 X3 X4 X5 X1*X2*X3 Any thoughts on how to resolve this? Thanks

jswislar · ‎01-05-2017

Thanks, that worked!

jswislar · ‎01-05-2017

Sure. Like this: Site_ID p_code Index 001 120 1 001 120 2 001 130 1 002 130 1 003 120 1 003 120 2 003 120 3 It's cases like site 002 that are giving me trouble.

jswislar · ‎01-05-2017

Yes, that gives me the count, but I want to it in a column so I can examine these cases in more detail and not have to rerun tabulate

Date Registered	‎11-17-2016 09:40 AM
Date Last Visited	‎03-16-2018 03:21 PM
Total Messages Posted	43

Online Status	Offline
Date Last Visited	‎03-16-2018 03:21 PM