About NonSleeper

NonSleeper · ‎04-16-2015

After running this program: data temp; set temp; where VAR1 ne . and VAR2 ne .; VAR3 = VAR1 - VAR2; run; the dataset temp only consists of observations where VAR1 and VAR2 have non-missing values. But my intention is just to generate VAR3 with valid values (i.e., if we include missing VAR1 or VAR2 in the calculation, the generated VAR3 will be missing as well). So as the title says, why does this program shrink the sample size and what is the correct alternative?

NonSleeper · ‎04-12-2015

Thanks for the advice. I did that and was able to create a format catalog within SASUSER. Some follow-up questions: 1. When I reload the permanent SAS file, it doesn't automatically "synchronize" with the format catalog in SASUSER. Does it mean I still need to apply formats on variables (e.g., like this: DATA HAVE; SET HAVE; FORMAT Var1 Var1Fmt.;) ? Is there some statement for the permanent data file to recognize the format catalog and align its variables with the defined formats? 2. It seems like a separate format file is still needed when sharing data, isn't it?

NonSleeper · ‎04-12-2015

Like the title says, I have a format library for variables' values but now have to re-apply the formats on the data every time using SAS (or restarting, to be more exact). I wonder if there's some way to apply the formats permanently on the data, so that we don't have to worry about that additional steps. For example, when sharing data, I don't really want to include both the data file and the format library in the package, but rather prefer to make it as compact as possible. I also encounter this problem: After applying formats for the data set in WORK library, I save it as a permanent data set. At that time, the saving process works fine, the output data works fine, other things are fine. So I close SAS and don't save the format library. When I reopen SAS, I fail to load the permanent data set, i.e., it comes to a blank table (although the file size is still ~100MB!). The log message is that "ERROR: Format ZFormat not found or couldn't be loaded for variable Z."

NonSleeper · ‎03-09-2015

My survey collected data sets at hierarchical locations that I now want to append them. They consist of the same variables for the most part, but there are some variables in this data set while not in others and vice versa, so I intend to use SET statement. My question arises because, when importing data from Excel, there are a few variables defined as numeric in some data sets and character in others, perhaps due to data entry process. This in turn makes the program not work. So is there a remedy to this problem?

NonSleeper · ‎03-07-2015

Thanks for all the help. I've not tried every suggested solution, but attempted to figure out one on my own and thought that a simple and potential solution may be: (1) Scan the response variable; (2) If the response contains "1", then variable "Listen_to_music" returns the value of "1". Otherwise, "Listen_to_music" is "0" (3) Use the similar code at step (2) for other categories. Would you think it's an OK approach and what would you suggest for a program to do so?

NonSleeper · ‎03-06-2015

In my data, there's question like: "What would you often do at weekends?". There are a few available answers and respondents can choose multiple options. But the questionnaire design step wasn't great, so one response looks like this "1. Listen to music; 3. Read novels; 6. Hang-out with friends" while another is like "2. Watch movies; 3. Read novels; 4. Play games". My question is whether there's a remedy which separates out this mixed answer and creates a set of dummy variables, so that the ouput looks like this: ID Music Movie Novel Game Work Friend 01 1 0 1 0 0 1 02 0 1 1 1 0 0 Any thoughts are appreciated.

NonSleeper · ‎03-06-2015

It's interesting that when I import the data from excel, SAS automatically converts the submission variable to date format. For example, the value of "26 February 2015 7:19 PM" becomes "26FEB15:19:19:49". I check the format and it's DATETIME16.; I guess the length is the reason the seconds shows up though they aren't in the original variable. I change the format to DATETIME14. and things look pretty good now.

NonSleeper · ‎03-04-2015

So my data comes from an online survey and consists of, among others, two variables: ID and Time of submission. The later is a string variable and looks like this: 2 February 2015 7:18 PM The question arises from the fact that some participants submitted their responses more than once. So I want to select the most recent submission and drop others. Can you suggest a macro-free program that can do that work? I've got stuck in figuring out a numeric measurement reflecting time of submission from the original variable or something like that which will help with sorting. Any thoughts are appreciated.

NonSleeper · ‎12-08-2014

I have several variables, let's say V1 to V3, each of them containing values of 1 or 0. How to write a statement for an array of these variables, so that a new variable (e.g., V4) will receive value of 1 if any among V1-V3 is 1; and 0 otherwise (i.e., all among V1-V3 are 0s)?

NonSleeper · ‎11-10-2014

The data looks like this: data temp; input ID Var1 Var2 Var3; datalines; 001 x11 x12 x13 002 x21 . x23 003 . . . ; I want to create a new variable (Count) which counts the frequency of valid values from Var1 to Var3, so the new data looks like: ID Var1 Var2 Var3 Count 001 x11 x12 x13 3 002 x21 . x23 2 003 . . . . ; I'm not sure if the count value for ID=003 is missing value or 0, but the idea is basically the same.

NonSleeper · ‎10-14-2013

The data look like this: data temp; input Var1 Var2; datalines; 5 . 7 . 8 . 2 1 3 1 1 2 1 4 4 5 6 5 6 7 6 8 ; run; The idea is: there are cases where values of Var1 are equal to values of Var2 elsewhere while the corresponding values of Var2 are missing (values 5, 7, 8). The other specifications are not important I think. How can I select values of Var2 as specified above (5, 7, 😎 and turn them into missing (or whatever value)? That way it looks like: 5 . 7 . 8 . 2 1 3 1 1 2 1 4 4 . 6 . 6 . 6 .

NonSleeper · ‎08-23-2013

Let me try state the issue more generally. Suppose the data looks like: data temp; input ID Var; 1 11 2 22 3 33 ...... (N-1) A N B ; run; So we have N observations and their values in this original order. There are N! ways to change (or permute) the ordering of these observation, e.g. consider this (call it permutation 1): 1 11 2 22 3 33 ...... N B (N-1) A The permutation 1 has only one change in the ordering of observations. I want to make the permutations with larger number of observations' reordering (e.g. 500 times). The proc plan above is able to produce all possible permutations but I just want to select some of them based on the number of reordering. Plus, if the number of observations is large (such as 1000) the number of permutations will be 'permutationally' huge. If you know another way to solve this issue, please advise. If you wonder what it's used for, the next step I will apply the ordering of the original data to the permuted ones. That is equivalent to random assignment of Var to IDs.

NonSleeper · ‎08-23-2013

I use proc plan to create permutation but feel not so satisfied with results. Consider data that have 1500 observations and I want to take 1000 permutations out of all possibilities. Here's the program I write: data permute1; proc plan; factors n=1000 obs=1500 perm / noprint; output out=permute2; run; data permute3; set permute2; set permute1 point=obs; run; What the permutation does is to do so one by one. That is, the first permutation changes the position of only one pair of variable (ID) and so on. Now I want to make it work a bit harder. E.g., for the above data, is there some way to force each permutation to have, let's say, 500 position changes?

NonSleeper · ‎08-22-2013

The data looks like this: data temp; input Group ID Var; datalines; 1 11 111 1 22 222 1 33 333 2 11 111 2 33 333 2 22 222 3 22 222 3 11 111 3 33 333 ; run; The set-up is: For each Group, there is a different ordering of ID with corresponding Var. That can be thought of as a result of a permutation process. Now I want to apply the order of ID in the first group to subsequent groups and don't change the order of Var. The new data should look like: 1 11 111 1 22 222 1 33 333 2 11 111 2 22 333 2 33 222 3 11 222 3 22 111 3 33 333 If possible, please give some solution other than sql or iml as I'm not familiar with them.

NonSleeper · ‎08-20-2013

Suppose the hypothetical data looks like this: data temp; input ID Rec1 OrderRec1 Rec2 OrderRec2; datalines; 001 10 1 . 1 001 . 2 10 2 001 11 3 11 3 001 12 4 . 4 001 . 5 12 5 ; The set-up is: For each cluster of ID (here I give only ID=001), there always exists pairs of Rec1 and Rec2 such that Rec1=Rec2. These variables can be non-missing for each observation (as in the third row) or one can be missing in one observation and appears in the other. The order variables are always available, they have certain categories (here 5) and indicate the ordering of Rec1 and Rec2. I now want to create a data that contrast Rec1 and Rec2 for each cluster of ID, and if they are equal, create a new variable (say OrderRec2New) that equals the value of OrderRec2. In other words, I want to make a new data with records that look like row 3. Unnecessary information from Rec2 and OrderRec2 can be deleted, it looks like this: 001 10 1 10 2 001 11 3 11 3 001 12 4 12 5

Online Status	Offline
Date Last Visited	‎05-12-2016 10:18 PM

How to not print the indicator variable in PROC SGPANEL?

Re: How to run statistics that exclude outliers/extreme values?

How to run statistics that exclude outliers/extreme values?

Re: Which of the following methods captures better estimate of populat...

Which of the following methods captures better estimate of population

Re: How to assign mean values of groups to missing observations?

How to assign mean values of groups to missing observations?

Re: How can I remove the modified template and go back to the default?

How can I remove the modified template and go back to the default?

How to assign values to the non-observed, or "legitimate skip"?

Re: How to run statistics that exclude outliers/extreme values?

Re: How to compare a set of variables, detect identical values, and re...

Re: How to compare a set of variables, detect identical values, and re...

Re: How to compare a set of variables, detect identical values, and re...

Re: How to assign values to the non-observed, or "legitimate skip"?

Re: How can I remove the modified template and go back to the default?

Re: How to assign mean values of groups to missing observations?

Why does this program shrink the sample size and what is the correct a...

Re: How to apply variables' values format permanently?

How to apply variables' values format permanently?

How to append data sets with variables defined as both character and n...

Re: How to creat a set of dummy variables from a rather mixed input?

How to creat a set of dummy variables from a rather mixed input?

Re: How to create a numeric measurement reflecting time in quite a mix...

How to create a numeric measurement reflecting time in quite a mixed f...

A basic question of array statement

How to count frequency of columns?

How to contrast values of two variables

Re: How to create (rather true) permutations

How to create (rather true) permutations

Data management question

Data management question