About cau83

cau83 · ‎08-16-2018

You would need to do it in a couple steps, eg: proc means data=yourdata; by group; var your-var; output out=group sum=total; proc means data=yourdata; by group subgroup; var your-var; output out=subgroup sum=subtotal; data pct_of_total; merge group subgroup; by group; pct_of_total=subtotal/total; run;

cau83 · ‎08-15-2018

Does it work in the DATA STEP / call execute without being in the import macro? Have you tried adding a put statement or something else so that you know whether your conditional execution is resolving the way you think it would?

cau83 · ‎08-15-2018

You can create a macro variable like this: %let year=2016; If you have all the years in one data set, you can combine this with a where statement like so: proc surveyfreq data-yourdata; where year=&year; other code; run; Unless there's a reason you need it to be separate you may be better off by using a BY statement instead though. Then you would just run the proc once, and get all of your output in one place. proc surveyfreq data=yourdata; BY year; other code; run;

cau83 · ‎08-15-2018

DATEPART() only requires one argument. What you're using with the format as the 2nd argument was part of the INPUT function in that other example. Just use: DOSdate=datepart(BR_DOSBEG);

cau83 · ‎08-10-2018

if index(var1,"With Defendants") > 0 then var2="With Defendants"; else var2="With Other Parties"; index() is a character function where the first argument is the string/column you want to search, and second argument is what you want to search for. If that second argument is found, the function returns a number equal to the starting position. Else, it returns 0. So the above looks to see if it contains defendants, and if not, assumes it contains parties. You could add other if else/else logic if necessary.

cau83 · ‎08-10-2018

Yes. If you wanted the next 12 months, make your window 13 instead of 12. Then, in a data step, do something like: FwdAvg=(FwdAvg*13 - Total)/12; This will temporarily get you the sum of the next 12 months + current month, then subtract current month and re-average.

cau83 · ‎08-10-2018

Do you have SAS/ETS? If so, PROC EXPAND will allow you to do this, it allows for many types of moving averages. The reverse transformation option will need to be combined with the desired moving average transformation option. From http://morgan.dartmouth.edu/Docs/sas92/support.sas.com/documentation/cdl/en/etsug/60372/HTML/default/etsug_expand_sect026.htm "You can compute a forward moving time window operation by combining a backward moving time window operator with the REVERSE operator. For example, the following statement computes a five-period forward moving average of X." convert x=y / transformout=( reverse movave 5 reverse );

cau83 · ‎08-10-2018

Have you figured out how to import a single file correctly? Then, can you modify the code to import two different files correctly? Then, we could help with the looping. Is your issue the looping or the reading in of data and formatting correctly? The former would be relatively easy to help with, the latter more difficult. I'd recommend you first go and read more about how to use a data step with the infile and input statements and take a crack at it.

cau83 · ‎08-08-2018

Before you do what I say below, read through the options for SAS logs. I am not aware of how to do this but it may be possible to reference it somehow. That being said, you can use this code to write your log out to a text file. Then you can read that back in and parse it. Read in each line as a record, in one giant column. Search for those phrases that precede run time numbers you want; only keep those records; and then use character functions to extract the times into a column that you can ultimately convert to numeric. PROC PRINTTO log="your-file.txt"; run;

cau83 · ‎08-08-2018

I added the code in CAPS below. Check the date first, then set count to zero if not that day. data want(drop=count); set have; by permno descending date; IF DATE = "29DEC2017"d THEN COUNT=1; ELSE if first.permno then count=0; count+1; if count>1; run;

cau83 · ‎08-08-2018

Visually I don't see a lot of difference b/w ZIP and ZINB-- I have not done extensive work where it ultimately mattered (only some exploratory analysis). Google will return a lot of information if you search for it. I am not familiar with the FMM proc. If t explicitly handles these distributions then it may be worth a shot. Regardless of PROC/method, the modeling ZI data should be a two part process as you describe, where it's estimating separately whether or not it's zero; and if not, then estimating the non-zero value. A similar concept exists with forecasting methods for intermittent count data (where zeroes often occur). I have more experience with this, but not in SAS as I do not have SAS Forecasting Server which is where those procedures live.

cau83 · ‎08-08-2018

For "creating a variable to use in the statement" create a macro variable. %let time1=YOUR-TIME; And then call it: &time1 < TIME < &time2

cau83 · ‎08-08-2018

You may want to use some kind of regression model suitable for zero-inflated data (ZIP or ZINB, where P and NB are Poisson or Negative Binomial distribution). Example using proc genmod here may be helpful https://stats.idre.ucla.edu/sas/dae/zero-inflatedpoisson-regression/

cau83 · ‎08-08-2018

Do you yet know how dates and times are stored in SAS? Date: # of days since 1/1/11960 Date time: # of secs since 1/1/1960 Time: Can be a date time, where you're only showing the time; commonly, this is just the of seconds since midnight and if you were to reformat as a date time your date would be 1/1/1960. I'd recommend that you convert your time to a # and make sure you know what is there. You can also do that with your thresholds. Use the SAS function hms like so: format time1 best8.; time1=hms(10,2,32);

cau83 · ‎08-08-2018

Let me preface this by saying I do not have particular expertise in either logistic regression or working with panel data. That being said... I am not aware of any technical limitation on using logit + panel data. PROC LOGISTIC has a CLASS statement, so you could use your date variable in that statement or just as a continuous independent variable. You should be okay with nominal/unordered data. Per SAS documentation For nominal response logistic models, where the possible responses have no natural ordering, the logit model can also be extended to a multinomial model known as a generalized or baseline-category logit model The part I'm least familiar with, this being panel data and potentially needing to track each person, I googled 'sas logit fixed effects' and found this post which mentions using PROC SURVEYLOGISTIC with the STRATA statement. That may be your best next stepl

Online Status	Offline
Date Last Visited	‎09-12-2018 12:21 PM

Re: Proc means calculate percent from total

Re: Trigger email on proc import

Re: Macro to call different years of data

Re: DateTime and DatePart

Re: Working with Strings

Re: 12 Month Forward Moving Average

Re: 12 Month Forward Moving Average

Re: .txt File Import into SAS

Re: output the timing of each data step into a table

Re: dropping last observations

Re: Multiplicative seasonality in PROC UCM

Re: PROC UCM forecast confidence interval: Why are the Level component...

Re: Working with Strings

Re: output the timing of each data step into a table

Re: how to select top 3 observations based on 2 conditions

Re: Trend function in SAS

Re: feature (lasso) selection and parameter's p-value

Re: Proc means calculate percent from total

Re: Trigger email on proc import

Re: Macro to call different years of data

Re: DateTime and DatePart

Re: Working with Strings

Re: 12 Month Forward Moving Average

Re: 12 Month Forward Moving Average

Re: .txt File Import into SAS

Re: output the timing of each data step into a table

Re: dropping last observations

Re: Two part model for healthcare costs

Re: SAS Time Range

Re: Two part model for healthcare costs

Re: SAS Time Range

Re: multinomial regression in panel data