Here is a link to my SAS blog webpage were you can find articles regarding short-term demand sensing, as well as other article.    Charlie Chase, Author at SAS Voices ... View more

you're welcome!   fyi - use the development branch - it's faster, and the only thing stopping it being merged is that the documentation isn't 100% updated. ... View more

@anilgvdbm wrote: Hi Experts,   I need a solution in if and then conditions for multiple variables.    I have data for 10 years and each year as one variable please find the example below Cutomer ID year1 year2 year3 year4 year5 year6 year7 year8 year9 year10 Tag 1 49 12 12 37 32 26 13 14 35 21   2 30 46 43 24 31 17 45 40 16 16   3 25 39 32 29 30 48 32 43 35 17   4 14 40 47 35 12 18 32 39 44 42   5 23 12 19 42 26 24 25 30 15 41     in the above table, i want to create tag as good, average and bad. how to use if and then for 10 variables at one go.   the conditions are if year1-year10 le 10 then Tag=good else if year1-year10 le 10 then Tag=average; else if  year1-year10 le 10 then Tag=poor;   Request you all to suggest me how to do in otherways.     Regards. Anil   Your example data should include at least on case each of "good", "average" and "poor". And you should indicate to us which observation or row of data is which. You rules in the if then are exactly the same for all three cases so makes little sense. By year1-year10 do you mean "subtract year10 from year1", "all values of the variables year1 through year10", "any value of the variables year1 through year10" or something else. It is not very clear from your code or narrative which is intended. ... View more

Four years is not very much data for smoothing. I think the best technique would be a simple one such as an average or a weighted average. If you use a weighted average, you probably want to assign more weight to the most recent data, such as w = {0.125, 0.25, 0.5, 1}; or standardize by defining w = w / sum(w); If you use this to develop a score for Customer_ID = 2, the score for that customer would be  Score = 0.125*40 + 0.25*24 + 0.5*14 + 1*26 = 23.47   For a reference, you can read about various kinds of moving averages and how to compute a moving average in SAS. But if you know that you have exactly four observations for each ID, you can use BY-groups processing and the FIRST.ID and LAST.ID variables to add the value to the data set. ... View more

For mappings, yo usually don't do loops (at least not in SAS). You can do table look-ups in several ways, including SQL join, and put function using a user defined format (created from your mappng table). ... View more

See the documentation. The code would likely be proc import datafile="\path\file.accdb" table="name_of_table" out=mylib.mydata dbms=access replace ; run; ... View more

If you post example data in a csv, you should also post the code you use to read that data into a SAS dataset. Only then can we recreate your dataset exactly as it is. ... View more

@anilgvdbm     You mentioned "I need to extract the common records from all 7 datasets"   Try using intersect aka set operators ... View more

OK . But you should post some more data which contains all the 12 month.   proc sql; create table REMOVE as  select *   from have    group by mobile_num     having sum(     Bucket_of_Audio_Heard    in  (" "       "-"     "less than 50%")   ) >=12 ; quit;   P.S. change something - remove dot in it and change variable name . ... View more

First off, don't write all in upper case - I will change the subject. Secondly why do you have 3000 datasets all the same, sounds like your running some procedure over data and outputting to separate datasets - this is not an efficient methodology and then leads to questions like yours.  Look up by group processing.  In your original data create a grouping variable, then do your procedure with the by group - it is faster, less coding, less disk space, and simpler to work with.  E.g. I want to do means on all groups of age 1-10, 11-20, 20+ data inter; set have; length group $20; if 1 <= age <= 10 then group="Group 1"; else if 11 <= age <= 20 then group="Group 2"; else group="Group 3"; run; proc means data=inter; by group; /* key line */ var weight; output out=want n=n mean=mean; run; This will process one dataset, by the given group, and produce one dataset with all the output needed. If you have another point, please give full details. ... View more

You seem averse to using macros? Why?    It’s the best method to import all at once. One other option, is to convert all to CSV - via a VBS script and then to import them all.    Here’s an example of how this can work, with first finding all the names of files and then importing them all. You’ll want to design a process that will ensure the structure of the data is the same across all datasets and what you want.   https://github.com/statgeek/SAS-Tutorials/blob/master/Import_all_files_one_type ... View more

Still does not fix the problem with the moving columns. The current format forces you to make guesses which data are where; this is ungood, as it may lead to wrong interpretations further down in the file. One could try to make those guesses, but I much prefer correctly structured data. Either have fixed columns (and, implicitly, a fixed record length!) or consistent column separators throughout.   Simply have your originator supply a record description with rules that cover all lines. In doing that, they will find their problems and correct them on their own, before the ruleset becomes larger than the dataset 😉 ... View more

The error message is saying that the macro that is being called is not defined. Is this code you wrote yourself?  Where is the code that defines the macro? ... View more

You can order your response variable (underwt, normal, overwt, obese).  A reasonable model would be a multinomial logistic model, fit with either PROC GENMOD or PROC LOGISTIC, using a cumulative logistic link.  See the documentation for examples (Ordinal Logistic Regression and Logistic Modeling with Categorical Predictors in the PROC LOGISTIC documentation, Ordinal Model for Multinomial Data in the PROC GENMOD documentation). Steve Denham ... View more