About iliyan

iliyan · ‎03-28-2018

Hello everyone, I am using the following sample code to generate my cure plots. proc genmod data=db.dir5lane2sort; class ident; MODEL tot=logaadtn loglength logiri/ dist=NEGBIN link=log; assess var=(logiri) /seed=603708000; run; However, as you are much well aware that this result is paired with simulated paths. However, I want to generate this something like: Is there a way to do it with PROC GEN MOD's options, or do I have to use another code?

iliyan · ‎03-22-2018

Hello Everyone, I have been working on generating regression models to fit collision data. The general model approach is to use Negative Binomial Regression, using a log link function. The model could have been fitted using PROC GENMOD, I understand that in that code using ODS Graphics it would be possible to CURE Plots. However, this would be at a cost of assuming a fixed dispersion parameter. As such NL MIXED was the way to go as seen below. Is there a way I can use NL MIXED to get my cure plots, or do I need to set the model up in PROC GENMOD? libname db 'V:\data'; /* General Model Form: TOT=exp(alpha1)(AADT)^beta1 (Length)^beta2 (IRI)^beta3 Dispersion (k)=exp(alpha2)(length)^beta4 */ proc nlmixed data=db.dir5lane2; parms alpha1 1 beta1 1 beta2 1 beta3 1 alpha2 1 beta4 1 ; eta = alpha1 + beta1*logaadtn + beta2*loglength + beta3*logiri; lambda = exp(eta); eta_k = alpha2+beta4*loglength; k = exp(eta_k); loglike = (lgamma(TOT+(1/k)) - lgamma(TOT+1) - lgamma(1/k) + TOT*log(k*lambda) - (TOT+(1/k))*log(1+k*lambda)); model TOT ~ general(loglike); run; Any suggestion are appreciated, Iliya

iliyan · ‎03-13-2018

Thank you Arthur, works like a charm. Have only used sas for 2 months, thank you for presenting this code as a sample.

iliyan · ‎03-12-2018

See attached updated file for the wanted result.

iliyan · ‎03-12-2018

Thank you for your once again for your quick reply Astounding! What I would like to select from the dir5lane2 dataset is all ident's that either not appears in the refactmin dataset or the ident does appear but the year is less than the refactmin specifies for that matching ident section. (dir5lane2.ident = refactmin.ident and dir5lane2.year < refactmin.year) or (dir5lane2.ident <> refactmin.ident) The dir5lane2 is only 16k observations, so there must be something with my understanding of proc sql. The code does run fine when I do not include the former restriction but only the latter.

iliyan · ‎03-12-2018

Hello Everyone, I am creating a database in which I include observations include segments that had no road work done and on also include any segment with road work done on it prior to the first time any work was done. However, when I try to match these conditions I get the following log response: NOTE: The "<>" operator is interpreted as "not equals". NOTE: The execution of this query involves performing one or more Cartesian product joins that can not be optimized. WARNING: Variable year already exists on file DB.REF1. WARNING: Variable ident already exists on file DB.REF1. ERROR: Insufficient space in file DB.REF1.DATA. ERROR: File DB.REF1.DATA is damaged. I/O processing did not complete. The code is as follows: proc sql; create table db.ref1 as select * from db.dir5lane2 road, db.refactmin act where (road.ident = act.ident and road.year < act.year) or act.ident <> road.ident; quit; As always there is a sample dataset attached. Any suggestion is appreciated.

iliyan · ‎03-10-2018

Well, that was "Astounding', that worked like a charm thank you very much.

iliyan · ‎03-10-2018

Hello Everyone, I am trying to convince observations into one like per data set (unique 'ident' label). The code here works, but for some reason, it does not do the final set of observations for year2013. However, if I shorten down the proc means var and sum ignoring some observations I have it works just fine. I know the code is not the most efficient but I have only been working with this program for 2 months, is there just a simple rule I am overlooking when using proc means? Also attached is a sample set of data for your reference. data db.ebcollector; set db.ebcollectorint; if year = 2000 then do; AADT_00 = AADT; iri_00 = iri; TOT_00 = TOT; FI_00 = FI; PD_00 = PD; end; if year = 2001 then do; AADT_01 = AADT; iri_01 = iri; TOT_01 = TOT; FI_01 = FI; PD_01 = PD; end; if year = 2002 then do; AADT_02 = AADT; iri_02 = iri; TOT_02 = TOT; FI_02 = FI; PD_02 = PD; end; if year = 2003 then do; AADT_03 = AADT; iri_03 = iri; TOT_03 = TOT; FI_03 = FI; PD_03 = PD; end; if year = 2004 then do; AADT_04 = AADT; iri_04 = iri; TOT_04 = TOT; FI_04 = FI; PD_04 = PD; end; if year = 2005 then do; AADT_05 = AADT; iri_05 = iri; TOT_05 = TOT; FI_05 = FI; PD_05 = PD; end; if year = 2006 then do; AADT_06 = AADT; iri_06 = iri; TOT_06 = TOT; FI_06 = FI; PD_06 = PD; end; if year = 2007 then do; AADT_07 = AADT; iri_07 = iri; TOT_07 = TOT; FI_07 = FI; PD_07 = PD; end; if year = 2008 then do; AADT_08 = AADT; iri_08 = iri; TOT_08 = TOT; FI_08 = FI; PD_08 = PD; end; if year = 2009 then do; AADT_09 = AADT; iri_09 = iri; TOT_09 = TOT; FI_09 = FI; PD_09 = PD; end; if year = 2010 then do; AADT_10 = AADT; iri_10 = iri; TOT_10 = TOT; FI_10 = FI; PD_10 = PD; end; if year = 2011 then do; AADT_11 = AADT; iri_11 = iri; TOT_11 = TOT; FI_11 = FI; PD_11 = PD; end; if year = 2012 then do; AADT_12 = AADT; iri_12 = iri; TOT_12 = TOT; FI_12 = FI; PD_12 = PD; end; if year = 2013 then do; AADT_13 = AADT; iri_13 = iri; TOT_13 = TOT; FI_13 = FI; PD_13 = PD; end; run; proc sort; by ident lengthn; run; proc means data=db.ebcollector noprint; by ident lengthn; var act AADT_00 iri_00 TOT_00 FI_00 PD_00 aadt_01 iri_01 tot_01 fi_01 pd_01 aadt_02 iri_02 tot_02 fi_02 pd_02 aadt_03 iri_03 tot_03 fi_03 pd_03 aadt_04 iri_04 tot_04 fi_04 pd_04 aadt_05 iri_05 tot_05 fi_05 pd_05 aadt_06 iri_06 tot_06 fi_06 pd_06 aadt_07 iri_07 tot_07 fi_07 pd_07 aadt_08 iri_08 tot_08 fi_08 pd_08 aadt_09 iri_08 tot_08 fi_08 pd_08 aadt_09 iri_09 tot_09 fi_09 pd_09 aadt_10 iri_10 tot_10 fi_10 pd_10 aadt_11 iri_11 tot_11 fi_11 pd_11 aadt_12 iri_12 tot_12 fi_12 pd_12 aadt_13 iri_13 tot_13 fi_13 pd_13; output out=db.ebcollectorsum sum=act AADT_00 iri_00 TOT_00 FI_00 PD_00 aadt_01 iri_01 tot_01 fi_01 pd_01 aadt_02 iri_02 tot_02 fi_02 pd_02 aadt_03 iri_03 tot_03 fi_03 pd_03 aadt_04 iri_04 tot_04 fi_04 pd_04 aadt_05 iri_05 tot_05 fi_05 pd_05 aadt_06 iri_06 tot_06 fi_06 pd_06 aadt_07 iri_07 tot_07 fi_07 pd_07 aadt_08 iri_08 tot_08 fi_08 pd_08 aadt_09 iri_08 tot_08 fi_08 pd_08 aadt_09 iri_09 tot_09 fi_09 pd_09 aadt_10 iri_10 tot_10 fi_10 pd_10 aadt_11 iri_11 tot_11 fi_11 pd_11 aadt_12 iri_12 tot_12 fi_12 pd_12 aadt_13 iri_13 tot_13 fi_13 pd_13; run;

iliyan · ‎02-13-2018

EDIT, The solution is: where mvab.dirtra = pave.direction and (pave.from_lhrs <= mvab.REFERENCE_NO < pave.to_lhrs); All that had to be done is to restrict the to_LHRS from matching.

iliyan · ‎02-13-2018

I want the lhrs in the mvab file to be bounded between the to and from lhrs of the pavement file.

iliyan · ‎02-13-2018

Hello everyone, I am trying to bring together two tables using the following code: proc sql; create table my.mvabpavement2008 as select * from my.mvab2008p mvab, my.pavement2008 pave where mvab.dirtra = pave.direction and mvab.REFERENCE_NO >= pave.from_lhrs and pave.to_lhrs >= mvab.REFERENCE_NO; quit; However, when running that code I get duplicate results especially in cases where REFERENCE_NO= from_lhrs and REFERENCE_NO (not)equal to_lhrs and the vice versa. I would think that this paraphrasing would indicate that unless statements are satisfied it would not produce an observation. What am I doing wrong? A sample is attached.

iliyan · ‎01-29-2018

Hello ballardw, thank you very much for your reply and interest in this matter. 1) Yes although I call them 1,2,3 and 4 those are not the original file names, the ones I use work no problem. But to ease confusion let us refer to them as sets: A, B, C, D and E(final). 2) In D I made a mistake I meant to indicate is as Count3 I am working on merging the files one at a time and not all at once. I will try the code. Frist in A,C and D ID/IDD is an identifier of a unique portion of a road. That portion is then defined by a continuous length identifier FROM and TO. In A there are two of the same IDs because another variable(not shown) is different the requiring to split it into 2 sections. In B only the IDD (where IDD=ID) and at some Off there is a point of interest. Thus the first line in B will only fall into the gap between 0 and 0.5 and not the second and vice versa for the second observation. Apart from the naming mistake, C and D are the same sections but the Count3 is the only observation and thus there are no splits.

iliyan · ‎01-29-2018

Hello everyone, I am currently working on a creating master file for further SAS Analysis. For the most part, all 4 files are formatted to produce the master file I desire but I cannot seem to bring the files together without losing data. For starters file 1 is the base and the other 3 files will add to it. The relevant portion of dataset 1 looks like so: ID From To 1 0 0.5 1 0.5 1 2 0 1 3 0.3 0.7 In dataset 2 is as such: IDD Off Count 1 0.2 10 1 0.7 5 2 0.5 8 3 0.3 20 Where IDD is the same as ID and Off in between From and To 3/4 are similar in concept and look like so: 3) ID FROM TO Count2 1 0 1 1.65 2 0 1 2.0 3 0.3 0.7 1.56 4) ID FROM TO Count3 1 0 1 18547 2 0 1 23506 3 0.3 0.7 6582 Where Count2 and Count are not related. I have tried the following code trying to merge 1 and 2 but the resultant (5) results in huge quantities of sections being lost (3000) observations to (700) observations. proc sql; create table my.5 as select * from my.1 as 1, my.2 as 2 where 1.ID = 2.IDD and 1.From<=2.Off and 1.To >= 2.Off; quit; Overall I wish the master file to look like: ID From To Count Count2 Count3 1 0 0.5 10 1.65 18547 1 0.5 1 5 1.65 18457 2 0 1 8 2.0 23506 3 0.3 0.7 20 1.56 6582 Any suggestions?

Online Status	Offline
Date Last Visited	‎03-28-2018 07:37 PM

Cumulative Residual Plots (CURE plots)

Cumulative Residual Plots (CURE plots)

Re: Proc SQL matching only missing values

Re: Proc SQL matching only missing values

Re: Proc SQL matching only missing values

Proc SQL matching only missing values

Re: Proc Means not generating result

Proc Means not generating result

Re: SQL PROC SORT producing duplicates

Re: SQL PROC SORT producing duplicates

Cumulative Residual Plots (CURE plots)

Cumulative Residual Plots (CURE plots)

Re: Proc SQL matching only missing values

Re: Proc SQL matching only missing values

Re: Proc SQL matching only missing values

Proc SQL matching only missing values

Re: Proc Means not generating result

Proc Means not generating result

Re: SQL PROC SORT producing duplicates

Re: SQL PROC SORT producing duplicates

SQL PROC SORT producing duplicates

Re: How to merge data with restrictions

How to merge data with restrictions