About gyambqt

gyambqt · ‎05-17-2017

Hi I think this is a very basic question. why I can't use %scan or some other macro function inside of a datastep. Is this because data step is reading value from PDV and %scan is creating a text? %let a=%str(a_b); data a; a= "'"||%scan(&a,2,_)||"'"; run; I wonder why the above program won't work. please provide detailed explaination and refer me to any document if it is relevant. Thanks

gyambqt · ‎05-16-2017

thanks for your reply but it doesn't work when you try %put %scan(%superq(a),2,%str(@));

gyambqt · ‎05-16-2017

Hello experts, if I have a string defined as &a and I want to grab all the string prior to @ from the whole string as defined by the %macro a %put. how do I achieve the result? the following macro a is not working... %let a=%str(SELECT a.ITS_ORIG_STN_CD , a.ITS_ORIG_STN_DS , b.CODE_CD, b.SUB_CODE_CD, b.CODE_TYPE_GP, b.CODE_LITERAL_DS from DS_LEAP.DS_ACCUSED a INNER JOIN DS_LEAP.DS_CODE_TABLE b ON ( a.ITS_ORIG_STN_CD = b.CODE_CD ) WHERE b.CODE_TYPE_GP = '0026' AND a.ITS_ORIG_STN_DS <> b.CODE_LITERAL_DS@a); 　 %macro a; 　 %put %scan(%nrstr(&a),1,'@'); 　 %mend; %a; Thanks

gyambqt · ‎05-16-2017

HI I modified my question. sorry for making it unclear

gyambqt · ‎05-15-2017

Hi Experts, I have a list of DI jobs stored in different metadata folder. I want to compare the name of the DI objects with metadata folder name. for example, I can have DI job called 001_job1 and 002_job2 and both jobs are stored as metadata object in metadata folder called 000_folder My question is how do I export or retrieve the name of each DI objects and it's corresponding metadata folder name? the output can be like: metadata_folder DI job 000_folder 001_job1 000_folder 002_job2 Thanks.

gyambqt · ‎05-10-2017

Also I found your code will check every single records. but if my data contain 100 millitons records it will perform a proc fre for all 100millions records and return a result which is not that efficent. Is there any way to improve the efficency?

gyambqt · ‎05-10-2017

Hi expert, if any of my table contain a mixture of character and numeric variable and I only want to test if any of the 8 columns has missing value. how do I do that? for example: I have 100 columns in one of my table A and 3 of the 8 columns are presented there and I want to test if the 3 columns are missing.

gyambqt · ‎05-09-2017

Hi expert, i have a base table contain 5 columns and I have appending table has 8 columns. the 5 columns are included within the 8 column. when I use proc append with force option. The 3 columns not included are disappearing. how do I append them using proc append and I can accept missing value after appending. thanks

gyambqt · ‎05-09-2017

I modified my question hope it is more clear. i build the code my self but I want to see if experts have better way of handling this problem. sorry can't share my code as I am typing using my phone.

gyambqt · ‎05-08-2017

yes, the dataset are consistent and always has the columns. however the column can be either numeric or date or character format.

gyambqt · ‎05-08-2017

sure, I will notice that in the future.

gyambqt · ‎05-08-2017

Hello Experts, I have got 100 SAS tables and they are all stored under same library. I want to check if some of the columns are populated correctly（do a missing value check,) there are about 8 columns I want to check however the 8 columns are not all presented in all the table. ，so some tables may have 2 out of the 8 columns，some may have 3 . all the 8 columns I want to check are in numeric value. The output table I want to have should have the following columns The table name that has been checked the library name the table stored column name that has at least one missing value I hope I made my question clear. THanks

gyambqt · ‎03-20-2017

cool, I have done things similar but takes ages to run for like 1.3 miliions applications. Maybe I need to find a way to make it more efficent processing. thanks a lot!!!!!

gyambqt · ‎03-20-2017

I realised using the following statment won't be able to find the correct result because It will ignore both from and to value if one of them is missing. for example: from to 1 missing then that record will be ignored. However it should be included as from=1 might connect to some applications if not missing(x{i}) and not missing(x{j}) then output; Also I think something need to be done at X table by giving a distinct number to all the missing value (so in other words, treat the missing value as distinct ID) so as to get the application number at the last step. anything wrong please let me know thank!

gyambqt · ‎03-19-2017

Thanks for your reply! your code is very nice! but to further extend a bit, if the data contain missing values like one or all the columns are blank for some applicaitons, how do you handle this problem? I modified your code a little like below: data have; set n; array x{*} $ Customer_Contact_Number name_dob_cluster1 driver_cluster1 address_75_cluster1; if driver_cluster1='1' then driver_cluster1=''; length from to $ 100; do i=1 to dim(x)-1; from=x{i}; do j=i+1 to dim(x); to=x{j}; 　 if from^='' or to^='' then output; end; end; keep from to; run; data have1; set have; if from='' then do; from1+1; from=cats('a',from1); end; if to='' then do; to1+1; to=cats('b',to1); end; run; so if all of the columns are blank then it should be removed. if one or some of the columns are blank then the blank columns will be assigned a ID like a1, a2 a3 (all the blank column must have a distinct id) what do you think?

Online Status	Offline
Date Last Visited	‎02-26-2024 02:16 AM

How to write a SAS program to find relationship between nodes

How to conduct impact analysis of a metadata table using SAS code

Re: Missing date column load data from Oracle to SAS

Missing date column load data from Oracle to SAS

Transfer text files from remote SFTP sever to local linux server

How to write a SAS code to extract SAS DI metadata objects?

How to identify log files differences between two log files?

Re: SAS bulkload issue

Re: SAS bulkload issue

Re: SAS bulkload issue

Re: why you can't use a macro funcation like %scan inside of a data st...

Re: How do I get the first part of the string using %scan

Re: How do I write a SAS program to re-construct the table?

Re: SAS bulkload issue

stored process questions (Stream vs Package outputs result)

why you can't use a macro funcation like %scan inside of a data step

Re: How do I get the first part of the string using %scan

How do I get the first part of the string using %scan

Re: How do output metadata folder and objects into a SAS table

How do output metadata folder and objects into a SAS table

Re: Identify columns with missing data - multiple datasets

Re: Identify columns with missing data - multiple datasets

How to append a table if it has more columns than base table used to a...

Re: Identify columns with missing data - multiple datasets

Re: Identify columns with missing data - multiple datasets

Re: Identify columns with missing data - multiple datasets

Identify columns with missing data - multiple datasets

Re: How to find link between nodes

Re: How to find link between nodes

Re: How to find link between nodes