About Kels123

Kels123 · ‎01-06-2022

Hello, I am doing a study looking at correlations between multiple cytokines. I found code to let me graph a heatmap of the correlation matrix, but I am having trouble making modifications to the heatmap. 1.) Specifically, I would like to be able to change the color, i.e. instead of white = -1 and dark blue = +1, I was hoping to use a different color for <1 values to make the correlation values more obvious. 2.) In addition, I was hoping to reorder the cytokines so that the strongest and weakest correlations are next to each other to make the data more interpretable. 3.) I also in a perfect world would be able to re-label the cytokine names so that show up prettier (e.g. with hyphens instead of underscores, etc). Here is my code below. The last paragraph of code is what I was trying to use to modify the order of the cytokines (labeled/* load the module for single-link clustering */). I am also attaching images of what my heatmaps currently look like. Any guidance would be greatly appreciated! /*HEAT MAP CORRELATION MATRIX FOR ALL SAMPLES*/ TITLE "Correlation Matrix for All Samples"; proc iml; /* create correlation matric for variables in alphabetical order */ varNames = {'IL_28A''IL_6''IL_8''IL_7''IFN_Beta''IL_10''CCL2''IL_1b''IFN_g''CCL3''CCL22''CCL4''IL_4' /*'IL_17A' 'IL_2'*/ 'CXCL9' 'IL_5' 'TNFRI' 'IL12p70' 'CCL17' 'TNFa' 'IL_28B' 'CXCL10' /*'IL_23'*/}; /*Sample_cat = sample_cat; Dz_severity_binary2 = Dz_severity_binary2;*/ use Gawlib.cytokine_80121 where (Dz_severity_binary2 >=0) ; read all var varNames into Y where (sample_cat>=1) ; read all var varNames into Y where (sample_cat>=1) ; close; corr = corr(Y, "Spearman"); call HeatmapCont(corr) xvalues=varNames yvalues=varNames colorramp=colors range={-1.01 1.01} title="Correlation Matrix: Cytokine levels for all samples"; /* load the module for single-link clustering */ load module=(SingleLinkCluster); permutation = SingleLinkCluster( corr ); /* permutation of 1:p */ V = varNames[permutation]; /* new order for variables */ R = corr[permutation, permutation]; /* new order for matrix */ call HeatmapCont(R) xvalues=V yvalues=V colorramp=colors range={-1.01 1.01} title="Correlation: Variables in Clustered Order";

Kels123 · ‎07-22-2021

Hello, I'm getting back into using SAS after a few years, and I am having trouble importing my excel file. I use SAS 64-bit on Windows 10 on Parallels 16 virtual environment, so I located the file location and the file name and used this code (which has worked for me previously): Proc import DATAFILE="C:\Users\Kelsey\Documents\UCSF COVID Cytokine Research Project - Gaw\cytokine_modified_7.1.21" OUT=Cytokine_72221 dbms=xls replace; run; However, I am getting this error message: ERROR: Physical file does not exist, C:\Users\Kelsey\Documents\UCSF COVID Cytokine Research Project - Gaw\cytokine_modified_7.1.21.xls. NOTE: The SAS System stopped processing this step because of errors. NOTE: PROCEDURE IMPORT used (Total process time): real time 0.03 seconds cpu time 0.03 seconds I have triple checked the file name and now I am concerned something else more complicated is going on. Please help! Thank you.

Kels123 · ‎07-18-2018

I put this code together using outputfmt (also tried imagefmt) but unfortunately PDF is not supported by the destination I am trying to save it to: ods graphics on / OUTPUTFMT=pdf imagename="K-M Curve CD4"; ods listing gpath="C:\Users\Kelsey\Desktop\SURVIVAL CURVES - PDF ATTEMPTS" ; proc lifetest data=work.tempCD4 method=km atrisk plots=(S(test nocensor atrisk (maxlen=16 outside(0.12))=0 to 3000 by 500) /*method=km MAXTIME= 3000 intervals=(0 to 3000 by 500)*/); time time*EVENT_ANYDEATH(0); strata CD4_atdiagnosis / order=internal adjust=sidak test=logrank ; label CD4_atdiagnosis = "CD4 count at diagnosis"; format CD4_atdiagnosis CD4_atdiagnosisfmt.; run; ods graphics off; Any additional suggestions for how to get around this? I find it hard to believe that there is no standardized way of saving a survival curve as a PDF directly from SAS. Thank you in advance.

Kels123 · ‎07-01-2018

Here is what we were instructed: "For any statistical images you should provide editable vector files (ie, the original artwork generated by the statistical package used to make the image); our preferred formats for these files are .ai, .eps, or .pdf." Do you think a .png file would be acceptable based on these guidelines? I was able to figure out how to generate a pdf file using the bolded statement. However, when looking at the PDF files produced, for some reason SAS is making 2 of the 3 the K-M lines formatted as dotted instead of simple solid lines. This is not how they show up in SAS, which is confusing me and I am not sure how to tweak the code to fix this. ods graphics on / OUTPUTFMT=pdf imagename="K-M Curve CD4"; ods listing gpath="J:\Kelsey data created 03.16.2016\Kelsey's mortality folder"; proc lifetest data=work.tempCD4 method=km atrisk plots=(S(test nocensor atrisk(maxlen=16 outside(0.12))=0 to 3000 by 500) /*method=km MAXTIME= 3000 intervals=(0 to 3000 by 500)*/); time time*EVENT_ANYDEATH(0); strata CD4_atdiagnosis / order=internal adjust=sidak test=logrank ; label CD4_atdiagnosis = "CD4 count at diagnosis"; format CD4_atdiagnosis CD4_atdiagnosisfmt.; run; ods graphics off;

Kels123 · ‎07-01-2018

Hi, I need to save my Kaplan-Meier curves in a format other that will be more easily editable for a journal. Right now, SAS is only allowing me to save the images as .png or .bmp files. Is there a way I can just add a outputfmt= statement to my code to output my curves as .ai, .eps, or .pdf files? Here is the code I am using below. Thank you in advance! data _null_; %let url = //support.sas.com/documentation/onlinedoc/stat/ex_code/142; infile "http:&url/templft.html" device=url; file 'macros.tmp'; retain pre 0; input; _infile_ = tranwrd(_infile_, '&', '&'); _infile_ = tranwrd(_infile_, '<' , '<'); if index(_infile_, '</pre>') then pre = 0; if pre then put _infile_; if index(_infile_, '<pre>') then pre = 1; run; %inc 'macros.tmp' / nosource; * Modify the template; %ProvideSurvivalMacros %let TitleText0 = "Kaplan-Meier Plot: Probability of death, stratified by CD4 count at time of HIV diagnosis"; %let yOptions = label="Survival Probability" linearopts=(viewmin=0.7 viewmax=1 tickvaluelist=(0.7 0.8 0.9 1.0)); %let xOptions = label="Time (days since release)" ; %let GraphOpts = DataContrastColors=(green red blue) DataColors=(green red blue); /*%let TitleText1 = &titletext0 "CD4 count at diagnosis " STRATUMID;*/ %CompileSurvivalTemplates *KM stratified by AIDS STAGE**************************************************************; proc format ; invalue CD4_atdiagnosiscat '>=500 cells/µL' = 1 '200-499 cells/µL' = 2 '<200 cells/µL' = 3; value CD4_atdiagnosisfmt 1 = '>=500 cells/µL' 2 = '200-499 cells/µL' 3 = '<200 cells/µL'; run; data tempCD4(drop=g); set CLEAN.databasesurv3(rename=(CD4_atdiagnosis=g)); CD4_atdiagnosis = input(g, CD4_atdiagnosiscat.); run; * Analysis; ods graphics on; proc lifetest data=work.tempCD4 method=km atrisk plots=(S(test nocensor atrisk(maxlen=16 outside(0.12))=0 to 3000 by 500) /*method=km MAXTIME= 3000 intervals=(0 to 3000 by 500)*/); time time*EVENT_ANYDEATH(0); strata CD4_atdiagnosis / order=internal adjust=sidak test=logrank ; label CD4_atdiagnosis = "CD4 count at diagnosis"; format CD4_atdiagnosis CD4_atdiagnosisfmt.; run; ods graphics off;

Kels123 · ‎07-01-2018

Hi SAS Community, I found a way to tailor my Kaplan-Meier plots to present the data I need to show, including number of subjects at risk. The only issue is that now I need to add the number censored in brackets next to the number at risk at each time point (every 500 days). I don't know how to modify the code to do this. The only option I have at the moment is the calculate the number censored by hand and add it in at the bottom using powerpoint, which is less than optimal due to formatting issues and potential for human error. Any help would be greatly appreciated. My code is below. Thank you. data _null_; %let url = //support.sas.com/documentation/onlinedoc/stat/ex_code/142; infile "http:&url/templft.html" device=url; file 'macros.tmp'; retain pre 0; input; _infile_ = tranwrd(_infile_, '&', '&'); _infile_ = tranwrd(_infile_, '<' , '<'); if index(_infile_, '</pre>') then pre = 0; if pre then put _infile_; if index(_infile_, '<pre>') then pre = 1; run; %inc 'macros.tmp' / nosource; * Modify the template; %ProvideSurvivalMacros %let TitleText0 = "Kaplan-Meier Plot: Probability of death, stratified by CD4 count at time of HIV diagnosis"; %let yOptions = label="Survival Probability" linearopts=(viewmin=0.7 viewmax=1 tickvaluelist=(0.7 0.8 0.9 1.0)); %let xOptions = label="Time (days since release)" ; %let GraphOpts = DataContrastColors=(green red blue) DataColors=(green red blue); /*%let TitleText1 = &titletext0 "CD4 count at diagnosis " STRATUMID;*/ %CompileSurvivalTemplates *KM stratified by AIDS STAGE**************************************************************; proc format ; invalue CD4_atdiagnosiscat '>=500 cells/µL' = 1 '200-499 cells/µL' = 2 '<200 cells/µL' = 3; value CD4_atdiagnosisfmt 1 = '>=500 cells/µL' 2 = '200-499 cells/µL' 3 = '<200 cells/µL'; run; data tempCD4(drop=g); set CLEAN.databasesurv3(rename=(CD4_atdiagnosis=g)); CD4_atdiagnosis = input(g, CD4_atdiagnosiscat.); run; * Analysis; ods graphics on; proc lifetest data=work.tempCD4 method=km atrisk plots=(S(test nocensor atrisk(maxlen=16 outside(0.12))=0 to 3000 by 500)); time time*EVENT_ANYDEATH(0); strata CD4_atdiagnosis / order=internal adjust=sidak test=logrank ; label CD4_atdiagnosis = "CD4 count at diagnosis"; format CD4_atdiagnosis CD4_atdiagnosisfmt.; run; ods graphics off;

Kels123 · ‎09-19-2017

I have a curve I am printing using proc lifetest: proc lifetest data=CLEAN.databasesurv3 plots=s outsurv=clean.SurvEst_AIDSSTAGE conftype=linear; time time * EVENT_ANYDEATH(0); strata disease_categ_dx2_num ; run; The problem is I need to do a few key modifications things: 1.) Modify the y-axis to start at 0.5 2.) Add number of individuals at risk (for each stratum if possible) 3.) Add the log-rank p-value in a legend (not as dire as the other two things) I have tried modifying the proc lifetest code, and also using proc gplot, proc template, and macros like %newsurv, but nothing has gotten close to what I need and I think now I am just thoroughly confused. I know this is a common problem, and I could really use a solution that doesn't require an in-depth expertise in programming. It seems like I should use "layout overlay", but that seems to be part of proc template, which is very intimidating at this point. This is the closest I have gotten so far, using this code: proc lifetest data=CLEAN.databasesurv3 method=km atrisk plots=(S(test nocensor atrisk=0 to 3000 by 500)) method=km MAXTIME= 3000 intervals=(0 to 3000 by 500); time time*EVENT_ANYDEATH(0); strata disease_categ_dx2_num / adjust=sidak test=logrank; /*format chemo_weeks_cat chemo_weeks_cat.;*/ run; I was thinking maybe I could combine it with something like this? Thank you in advance. %ProvideSurvivalMacros %let yOptions = label="Survival Probability" linearopts=(viewmin=0.6 viewmax=1 /*tickvaluelist=(0 .2 .4 .6 .8 1.0))*/; %CompileSurvivalTemplates Any assistance would be greatly appreciated. If I DO need to use proc template or a macro, it would be really helpful to have a better explanation of how these work together with proc lifetest than the explanations I have found previously online. Thank you in advance.

Kels123 · ‎09-08-2017

I agree that this identifier is sub-optimal, but unfortunately I need to do a datamatch with a database that only has this identifier available. One quick follow-up question: How do I remove the slashes from the identifier? I am currently using this code: data match.mydatabase3; set match.mydatabase3; length URN $100; URN=upcase(cats(substr(first_name,1,1),substr(first_name,3,1),substr(last_name,1,1),substr(last_name,3,1),put(dob,MMDDYY8.),put(Gender_forTLCmatch,1.),"U")); run; Which produces this: SRSI09/01/992U Instead of this (which is what I want): SRSI0901992U Thanks very much.

Kels123 · ‎09-08-2017

I would like to create a variable from first name, last name, and date of birth that looks like this: First letter of first name + 3rd letter of first name + first letter of last name + 3rd letter of last name + DOB (mmddyy) + indicator of race (1-4) + Letter U For example, for a person with this data: First_name Last_name DOB Race Sarah Smith 09/01/1999 2 I want to create a variable that looks like this: SRSI0901992U Any help would be greatly appreciated. Thank you in advance.

Kels123 · ‎05-22-2017

Through a data request, I am working with a database of dates and corresponding measurements. In order to use the dates, I want to make sure they are all in the same format and readable/usable as dates but I am running into some unexpected issues. The date variables are called insurance_dt1 through insurance_dt26. Through proc contents, I can see that all the date variables are numeric with a length of 8., but only 1-15 have an associated date format (YYMMDD10.) whereas 16-26 have no date formatting. Looking closer, I can see why-- some of the date variables were provided to me in a different format that I do not immediately recognize. For example: insurance_dt14 insurance_dt15 insurance_dt16 insurance_dt17 insurance_dt18 insurance_dt19 2013-05-21 2013-06-05 19740 20355 19358 19390 I have tried converting them into YYMMDD10. format using the following codes within a datastep, but they usually just turn into missing values or the date 1960-01-02 or 1960-01-03 : new_ins_dt16=input(ins_dt16,yymmdd10.); format new_ins_dt16 yymmdd10.; new_ins_dt16 = INPUT(ins_dt16,8.); FORMAT new_ins_dt16 YYMMDD10.; new_ins_dt16 = INPUT(PUT(ins_dt16,8.),YYMMDD8.); FORMAT new_ins_dt16 YYMMDD10.; I am wondering if I am not properly recognizing the format and thus somehow causing SAS to be unable to read in the dates properly. Alternatively, these dates could be partially incomplete (e.g. missing the exact day), but that does not seem right either since these days visually appear to be more like unformatted SAS dates as opposed to incomplete days. I am not sure, for example, how to interpret something like "20355". Any suggestions would be very much appreciated. Thank you.

Kels123 · ‎05-02-2017

Hi, I have a dataset with about 1,300 subjects, many of which have multiple rows because they have multiple observation periods during my eight-year data collection / follow-up period. data tempfile; infile datalines truncover; INPUT StudyID$2.@+1 count StartDate mmddyy10.@+1 EndDate1 mmddyy10.@+1; format Startdate EndDate1 mmddyy10.; datalines; 01 1 02/01/2007 05/01/2008 02 1 06/02/2012 05/24/2013 03 1 05/07/2010 08/19/2010 04 1 06/16/2007 06/21/2008 05 1 04/28/2007 11/07/2008 06 1 03/16/2007 06/20/2007 07 1 04/18/2007 04/19/2007 08 1 05/11/2007 08/03/2007 09 1 08/01/2007 08/28/2007 09 3 02/21/2008 03/05/2008 10 1 02/06/2008 03/12/2010 11 1 01/31/2007 12/26/2008 11 3 02/11/2009 05/08/2009 11 5 05/19/2009 08/14/2009 11 7 09/22/2009 12/18/2009 11 9 01/09/2010 02/03/2010 11 11 02/23/2010 11/19/2010 11 13 12/01/2010 03/18/2011 11 15 05/10/2011 04/23/2013 11 17 08/03/2013 09/27/2013 ; run; proc print data=tempfile; run; Please note: The count variable is an old variable left over from a previous coding procedure. For subjects with more than one observation, I will need to choose which observation to use as my "starting point". In an effort to avoid bias by only looking at everyone's first period or looking at everyone's last period that would allow for adequate follow-up time, I was thinking I could randomly select which observation period to use for analysis. This would mean that I need to include everyone who has only one row and then, for subjects with more than one row, randomly selecting one of their rows to include for analysis. Is there a way to randomly assign ones and zeroes for example, in order to accomplish this? For example: StudyID count StartDate EndDate1 INCLUDE 01 1 02/01/2007 05/01/2008 1 02 1 06/02/2012 05/24/2013 1 03 1 05/07/2010 08/19/2010 1 04 1 06/16/2007 06/21/2008 1 05 1 04/28/2007 11/07/2008 1 06 1 03/16/2007 06/20/2007 1 07 1 04/18/2007 04/19/2007 1 08 1 05/11/2007 08/03/2007 1 09 1 08/01/2007 08/28/2007 1 09 3 02/21/2008 03/05/2008 0 10 1 02/06/2008 03/12/2010 1 11 1 01/31/2007 12/26/2008 0 11 3 02/11/2009 05/08/2009 0 11 5 05/19/2009 08/14/2009 0 11 7 09/22/2009 12/18/2009 1 11 9 01/09/2010 02/03/2010 0 11 11 02/23/2010 11/19/2010 0 11 13 12/01/2010 03/18/2011 0 11 15 05/10/2011 04/23/2013 0 11 17 08/03/2013 09/27/2013 0 Thank you very much in advance!

Kels123 · ‎04-27-2017

I have a database where subjects have repeated measurements. These measurements are taken during specific time periods delineated by StartDate and EndDate1. DAYSBWN1 is the length of time of each period (EndDate1 – StartDate in each row). Sample code below: data tempfile; infile datalines truncover; INPUT StudyID$2.@+1 StartDate mmddyy10.@+1 EndDate1 mmddyy10.@+1 daysbwn1 ; format Startdate EndDate1 mmddyy10.; datalines; 01 02/01/2007 05/01/2008 455 02 06/02/2012 05/24/2013 356 03 05/07/2010 08/19/2010 104 04 06/16/2007 06/21/2008 371 05 04/28/2007 11/07/2008 559 06 03/16/2007 06/20/2007 96 07 04/18/2007 04/19/2007 1 08 05/11/2007 08/03/2007 84 09 08/01/2007 08/28/2007 27 09 02/21/2008 03/05/2008 13 10 02/06/2008 03/12/2010 765 11 01/31/2007 12/26/2008 695 11 02/11/2009 05/08/2009 86 11 05/19/2009 08/14/2009 87 11 09/22/2009 12/18/2009 87 11 01/09/2010 02/03/2010 25 11 02/23/2010 11/19/2010 269 11 12/01/2010 03/18/2011 107 11 05/10/2011 04/23/2013 714 11 08/03/2013 09/27/2013 55 ; run; proc print data=tempfile; run; I would like to write a code that creates two new variables: DAYSBWN_CUM: The total number of time for all follow-up periods aka repeat rows (DAYSBWN1 added up for any rows that follow). REPEAT_NUM: The total number of times the subject has a repeat row (REPEAT_NUM is > 1 only if the subject has more than one row). The goal would be to get something that looks like this: StudyID StartDate EndDate1 daysbwn1 DAYSBWN_CUM REPEAT_NUM 01 02/01/2007 05/01/2008 455 0 0 02 06/02/2012 05/24/2013 356 0 0 03 05/07/2010 08/19/2010 104 0 0 04 06/16/2007 06/21/2008 371 0 0 05 04/28/2007 11/07/2008 559 0 0 06 03/16/2007 06/20/2007 96 0 0 07 04/18/2007 04/19/2007 1 0 0 08 05/11/2007 08/03/2007 84 0 0 09 08/01/2007 08/28/2007 27 13 1 09 02/21/2008 03/05/2008 13 0 0 10 02/06/2008 03/12/2010 765 0 0 11 01/31/2007 12/26/2008 695 1430 8 11 02/11/2009 05/08/2009 86 1344 7 11 05/19/2009 08/14/2009 87 1257 6 11 09/22/2009 12/18/2009 87 1170 5 11 01/09/2010 02/03/2010 25 1145 4 11 02/23/2010 11/19/2010 269 876 3 11 12/01/2010 03/18/2011 107 769 2 11 05/10/2011 04/23/2013 714 55 1 11 08/03/2013 09/27/2013 55 0 0 If the coding would be significantly easier, I could work with something like this instead, because I could just subtract certain values from the rows to get my desired final product: StudyID StartDate EndDate1 daysbwn1 DAYSBWN_CUM REPEAT_NUM 01 02/01/2007 05/01/2008 455 455 1 02 06/02/2012 05/24/2013 356 356 1 03 05/07/2010 08/19/2010 104 104 1 04 06/16/2007 06/21/2008 371 371 1 05 04/28/2007 11/07/2008 559 559 1 06 03/16/2007 06/20/2007 96 96 1 07 04/18/2007 04/19/2007 1 1 1 08 05/11/2007 08/03/2007 84 84 1 09 08/01/2007 08/28/2007 27 40 2 09 02/21/2008 03/05/2008 13 13 1 10 02/06/2008 03/12/2010 765 765 1 11 01/31/2007 12/26/2008 695 2125 9 11 02/11/2009 05/08/2009 86 1430 8 11 05/19/2009 08/14/2009 87 1344 7 11 09/22/2009 12/18/2009 87 1257 6 11 01/09/2010 02/03/2010 25 1170 5 11 02/23/2010 11/19/2010 269 1145 4 11 12/01/2010 03/18/2011 107 876 3 11 05/10/2011 04/23/2013 714 769 2 11 08/03/2013 09/27/2013 55 55 1 Any assistance would be greatly appreciated. Thank you in advance!

Kels123 · ‎04-25-2017

I am trying to use PROC SQL left join to merge two databases. The first database (master) is my actual database which has rows indicating time periods during which measurements were taken for each subject; I want to keep ALL of those rows. I want to add in data from the second database (additional) based on whether "additional" has values that correspond to the time periods in each of the rows in "master". If there are no measurements in “additional” that correspond to the specified time periods, I would like for those rows to be retained as missing values, instead of dropped. For instance, for example Subject #001, I want to merge data from “additional” into the “master” database to produce the “want” database below: Master: StudyID StartDate_C EndDate1_C Date_12mopost 001 01/04/2012 07/10/2012 01/04/2013 001 03/07/2013 06/18/2013 03/07/2014 001 06/25/2013 01/01/2015 06/24/2014 Want (desired combined database): StudyID StartDate_C EndDate1_C Date_12mopost full_lab_Date 001 01/04/2012 07/10/2012 01/04/2013 . 001 03/07/2013 06/18/2013 03/07/2014 . 001 06/25/2013 01/01/2015 06/24/2014 07/31/2013 001 06/25/2013 01/01/2015 06/24/2014 11/22/2013 I used left join (code below), but it looks like this code is still causing rows without matches in “additional” to be dropped: proc sql; create table want as select c.*, v.* from WORK.master AS c left join WORK.additional AS v on c.StudyID=v.StudyID_copy where c.StartDate_C <= v.full_lab_date <= c.DATE_12MOPOST ; ; quit; want (using the code above, which produces something different than what I want): StudyID StartDate_C EndDate1_C Date_12mopost full_lab_DATE 001 06/25/2013 01/01/2015 06/24/2014 07/31/2013 001 06/25/2013 01/01/2015 06/24/2014 11/22/2013 Is there a modification I can make to my code in order to get the desired outcome? Thank you in advance.

Kels123 · ‎04-24-2017

Hello, I have a database where subjects have repeated measurements. These measurements are taken during specific time periods delineated by StartDate and EndDate1. Sample code below: data tempfile; infile datalines truncover; INPUT StudyID$2.@+1 count StartDate mmddyy10.@+1 EndDate1 mmddyy10.@+1; format Startdate EndDate1 mmddyy10.; datalines; 01 1 02/01/2007 05/01/2008 02 1 06/02/2012 05/24/2013 03 1 05/07/2010 08/19/2010 04 1 06/16/2007 06/21/2008 05 1 04/28/2007 11/07/2008 06 1 03/16/2007 06/20/2007 07 1 04/18/2007 04/19/2007 08 1 05/11/2007 08/03/2007 09 1 08/01/2007 08/28/2007 09 3 02/21/2008 03/05/2008 10 1 02/06/2008 03/12/2010 11 1 01/31/2007 12/26/2008 11 3 02/11/2009 05/08/2009 11 5 05/19/2009 08/14/2009 11 7 09/22/2009 12/18/2009 11 9 01/09/2010 02/03/2010 11 11 02/23/2010 11/19/2010 11 13 12/01/2010 03/18/2011 11 15 05/10/2011 04/23/2013 11 17 08/03/2013 09/27/2013 ; run; proc print data=tempfile; run; Please note that the count variable is left over from a previous data manipulation, which is why there are only odd values instead of sequential numbers (1,3,5,... instead of 1,2,3,...). I want to further group these measurements into 12-month time periods that start on EndDate1; basically I am interested in following individuals for 12 months after EndDate1. The issue is that I am not sure how to write a code that will allow me to reorganize the data this way. I was thinking about trying to create something like this: StudyID count StartDate EndDate1 Date_12mopost Period 01 1 02/01/2007 05/01/2008 2009-05-01 1 02 1 06/02/2012 05/24/2013 2014-05-24 1 03 1 05/07/2010 08/19/2010 2011-08-19 1 04 1 06/16/2007 06/21/2008 2009-06-21 1 05 1 04/28/2007 11/07/2008 2009-11-07 1 06 1 03/16/2007 06/20/2007 2008-06-19 1 07 1 04/18/2007 04/19/2007 2008-04-18 1 08 1 05/11/2007 08/03/2007 2008-08-02 1 09 1 08/01/2007 08/28/2007 2008-08-27 1 09 3 02/21/2008 03/05/2008 2008-08-27 1 10 1 02/06/2008 03/12/2010 2011-03-12 1 11 1 01/31/2007 12/26/2008 2009-12-26 1 11 3 02/11/2009 05/08/2009 2009-12-26 1 11 5 05/19/2009 08/14/2009 2009-12-26 1 11 7 09/22/2009 12/18/2009 2009-12-26 1 11 9 01/09/2010 02/03/2010 2011-02-03 2 (starting over) 11 11 02/23/2010 11/19/2010 2011-02-03 2 11 13 12/01/2010 03/18/2011 2012-03-18 3 (starting over) 11 15 05/10/2011 04/23/2013 2014-04-23 4 (starting over) 11 17 08/03/2013 09/27/2013 2014-04-23 4 This will allow me to do what I need to do with the data to eventually be able to analyze it (for example, eventually transposing the data into the appropriate one-year time period, or selecting the first row of the one-year time period). So I guess my questions might be: How do I assign a date 12 months after the EndDate1 IN THE FIRST ROW FOR EACH SUBJECT (e.g. Subject 09)? How about later on, for repeat measurements, if the EndDate1 takes place AFTER the 12-month mark, and I want to start the clock over again with a new 12 month period (e.g. Subject 11)? Can a create a new “Period” variable to reflect these dates, to make it easier to keep the 12-month time periods organized? So far, I have only managed to create a semi-accurate date variable representing 12 months after EndDate1, using the following code: data tempfile; set tempfile; by studyID count; retain Date12mopost; if first.studyID then do; Date_12mopost=(EndDate1+365); end; else do; Date_12mopost=Date_12mopost; end; format Date_12mopost yymmdd10. ; run; proc print data = tempfile; run; StudyID count StartDate EndDate1 Date_12mopost 01 1 02/01/2007 05/01/2008 2009-05-01 02 1 06/02/2012 05/24/2013 2014-05-24 03 1 05/07/2010 08/19/2010 2011-08-19 04 1 06/16/2007 06/21/2008 2009-06-21 05 1 04/28/2007 11/07/2008 2009-11-07 06 1 03/16/2007 06/20/2007 2008-06-19 07 1 04/18/2007 04/19/2007 2008-04-18 08 1 05/11/2007 08/03/2007 2008-08-02 09 1 08/01/2007 08/28/2007 2008-08-27 09 3 02/21/2008 03/05/2008 2009-03-05 10 1 02/06/2008 03/12/2010 2011-03-12 11 1 01/31/2007 12/26/2008 2009-12-26 11 3 02/11/2009 05/08/2009 2010-05-08 11 5 05/19/2009 08/14/2009 2010-08-14 11 7 09/22/2009 12/18/2009 2010-12-18 11 9 01/09/2010 02/03/2010 2011-02-03 11 11 02/23/2010 11/19/2010 2011-11-19 11 13 12/01/2010 03/18/2011 2012-03-17 11 15 05/10/2011 04/23/2013 2014-04-23 11 17 08/03/2013 09/27/2013 2014-09-27 The Date_12mopost is accurate for each row, but doesn’t help me try to group the periods into the a 12-month time period based on the dates from the preceding row. I recognize this is a multi-step question that is likely extremely complicated. Any guidance, particularly regarding the specifics of the code, would be greatly appreciated. Thank you very much in advance!

Kels123 · ‎03-10-2017

Hello, I have written a code to plot a survival curve using PROC LIFETEST: proc lifetest data=clean.DATABASE3 plots=survival (FAILURE atrisk (atrisktick maxlen=13 outside)=0 14 30 90 180 365) ; time time1_C*censor1_C(1) ; run; quit; Because I want to better visualize my time points of particular interest, which are before the 365-day mark, I want to modify my x-axis to end at 365 days. However, I don't know how to do this. I tried using the overlay statement, but I don't think I understand how it is supposed to be used: proc template; layout overlay / xaxisopts=(shortlabel=XNAME offsetmin=.05 linearopts=(viewmax=365 tickvaluelist=XTICKVALS tickvaluefitpolicy=XTICKVALFITPOL); endlayout; run; Please help. I just wanted to know the simplest way to modify the x-axis when using PROC LIFETEST. Thanks in advance.

Online Status	Offline
Date Last Visited	‎01-11-2022 03:40 PM

How to modify a correlation matrix heatmap in PROC IML

Data import error: Physical file does not exist

Re: How do I save a Kaplan-Meier curve as .ai, .eps, or .pdf file?

Re: How do I save a Kaplan-Meier curve as .ai, .eps, or .pdf file?

How do I save a Kaplan-Meier curve as .ai, .eps, or .pdf file?

Kaplan-Meier Curve: How to show number censored below plot?

Modifications to Kaplan-Meier Curve for printing: y-axis and number at...

Re: How to create an identifier using first name, last name, and date ...

How to create an identifier using first name, last name, and date of b...

How to read and reconcile dates given in two different formats

Re: How PROC SQL helps you see missing values and other data attribute...

Re: How to combine certain select rows (but not all rows) into single ...

Re: How to select the max value from a range of measurements over time

Re: How do I combine time periods based on date and environment?

Re: Working with incomplete dates

How to modify a correlation matrix heatmap in PROC IML

Data import error: Physical file does not exist

Re: How do I save a Kaplan-Meier curve as .ai, .eps, or .pdf file?

Re: How do I save a Kaplan-Meier curve as .ai, .eps, or .pdf file?

How do I save a Kaplan-Meier curve as .ai, .eps, or .pdf file?

Kaplan-Meier Curve: How to show number censored below plot?

Modifications to Kaplan-Meier Curve for printing: y-axis and number at...

Re: How to create an identifier using first name, last name, and date ...

How to create an identifier using first name, last name, and date of b...

How to read and reconcile dates given in two different formats

How to randomly select at least one observation per subject

How to get a cumulative sum of all subsequent rows

How PROC SQL helps you see missing values and other data attributes

How to re-group/organize row observations based on dates in certain ro...

Modifying x-axis range in PROC LIFETEST