topic monthly data format in SAS Studio

monthly data format

SystemsasdD2019 — Wed, 08 Jan 2020 20:49:29 GMT

please , I want to change format my data, monthly,I Imprt it from stata program , start from 2016m1 , 2016m2 , 2016m3 ..... . to 2019m12 . ,It is 48 value , the date variable name, "month" the data is gold price and silver price my application on VARMAX on SAS on deamond academic . I need your help . with many thanks ,

Re: monthly data format

PaigeMiller — Wed, 08 Jan 2020 20:56:20 GMT

It would help us greatly if you could format your question in a readable way.

You also say you want to change the format of your data, but you don't show any data and you don't say what the end result should be. We need to know these things in order to be able to assist.

Re: monthly data format

Reeza — Wed, 08 Jan 2020 21:22:36 GMT

Please show what you have as input, what you want as output and anything you've tried so far. Please post your data using the code blocks and use full sentences whenever possible.

Re: monthly data format

PGStats — Wed, 08 Jan 2020 21:56:35 GMT

As far as I know, SAS doesn't provide an informat to read such fields directly. If you are looking for a way to read dates in the YYYYmMM format, you would have to use something like:

data have;
str = "2019m2";
date = input (cats(translate(str,"-","m"),"-1"), yymmdd10.);
format date yymmdd10.;
run;

proc print; run;

Re: monthly data format

Kurt_Bremser — Thu, 09 Jan 2020 06:50:43 GMT

If I interpret your (very confusing) post correctly, you should look at this:

data have;
input inmonth $7.;
datalines;
2016m1
2016m2
2019m12
;

data want;
set have;
outmonth = input(compress(inmonth,'m'),yymmn6.);
format outmonth yymm7.;
run;

Re: monthly data format

SystemsasdD2019 — Thu, 09 Jan 2020 10:24:27 GMT

thanks very much for your answer, but I think the data is unclear , I will send my details to be clear to you

Re: monthly data format

SystemsasdD2019 — Thu, 09 Jan 2020 13:30:45 GMT

thank you for your answar , I have implemented the program and the result is shown below , 


data have;
input inmonth $7.;
datalines;
2016m1
2016m2
2016m3
2016m4
2016m5
2016m6
2016m7
2016m8
2016m9
2016m10
2016m11
2016m12
2017m1
2017m2
2017m3
2017m4
2017m5
2017m6
2017m7
2017m8
2017m9
2017m10
2017m11
2017m12
2018m1
2018m2
2018m3
2018m4
2018m5
2018m6
2018m7
2018m8
2018m9
2018m10
2018m11
2018m12
2019m1
2019m2
2019m3
2019m4
2019m5
2019m6
2019m7
2019m8
2019m9
2019m10
2019m11
2019m12
;

data want;
set have;
outmonth = input(compress(inmonth,'m'),yymmn6.);
format outmonth yymm7.;
run;

but when I implemented the program :

PROC VARMAX DATA=WORK.IMPORT PRINTALL PLOTS=ALL;
MODEL LSILVERPRICE LGOLDPRICE/DIF=(LSILVERPRICE(1) LGOLDPRICE(1)) Q=1 METHOD=ML;
OUTPUT LEAD=25;
ID MONTH INTERVAL=MONTH;
RUN;



An error appears :
ERROR : Variable month in list does not match type prescribd for this list

Re: monthly data format

Kurt_Bremser — Thu, 09 Jan 2020 13:57:33 GMT

I guess that your variable month is of type character.

From the documentation of the VARMAX Procedure:

The ID statement specifies a variable that identifies observations in the input data set. The datetime variable specified in the ID statement is included in the OUT= data set if the OUTPUT statement is specified. The ID variable is usually a SAS datetime variable. The values of the ID variable are extrapolated for the forecast observations based on the value of the INTERVAL= option.

(emphasis by me)

The ID variable should therefore be numeric; it might be that a date value will also work, but to be sure, you should convert it to a datetime value (count of seconds from 1960-01-01T00:00:00)

Note that my code creates outmonth as a numeric date variable, which can be converted to datetime by using

dhms(outmonth,0,0,0)

Re: monthly data format

SystemsasdD2019 — Thu, 09 Jan 2020 15:50:17 GMT

Sorry , I am a beginner in the SAS program , so how do I implement the conversion from numerical to datetime sas.

Re: monthly data format

PGStats — Thu, 09 Jan 2020 18:41:25 GMT

Change your program to:

data want;
set have;
outmonth = dhms(input(compress(inmonth,'m'),yymmn6.), 0,0,0);
format outmonth datetime.;
run;

Re: monthly data format

SystemsasdD2019 — Thu, 09 Jan 2020 22:02:42 GMT

thank you very mach for your answer , but when I implement the program

PROC VARMAX DATA=WORK.IMPORT PRINTALL PLOTS=ALL;
MODEL LSILVERPRICE LGOLDPRICE/DIF=(LSILVERPRICE(1) LGOLDPRICE(1)) Q=1 METHOD=ML;
OUTPUT LEAD=25;
ID outmonth INTERVAL=month;
RUN;

An error appears :
ERROR: Variable outmonth not found .

Re: monthly data format

Kurt_Bremser — Thu, 09 Jan 2020 22:11:07 GMT

You need to adapt the code to your data. Use variables that exist in your dataset. If you created outmonth in a new dataset, use this dataset in the varmax procedure.

Re: monthly data format

PGStats — Fri, 10 Jan 2020 04:21:09 GMT

You should refer to the transformed data, not the original version.

Re: monthly data format

SystemsasdD2019 — Fri, 10 Jan 2020 16:10:25 GMT

Sory ,my orijinal data (goldprice and silverprice) in table WORK.IMPORT and OUTMONTH created in table WORK.WANT how can I put it them in one table , So I can apply my program PROC VARMA .

Re: monthly data format

SystemsasdD2019 — Sun, 12 Jan 2020 21:01:22 GMT

Sory , when I implement this program :An error apear

proc varmax data=WORK.WANT plot(unpack)=(residual model forecasts);
id OUTMONTH interval=month;
model Lsilverprice Lgoldprice  / dify=(1,12) noint p=0 q=1;
output lead=12;
run;

An ERROR : variable Lsilverprice not found

ERROR variable Lgoldprice not found

I refer to data set of variable of date OUTMONTH but the dataset of variabls Lsilverprice Lgoldprice is WORK.IMPORT . how can do my code with many thanks .

Re: monthly data format

Kurt_Bremser — Sun, 12 Jan 2020 21:52:40 GMT

Maxim 3: Know Your Data. You can only run the procedure on variables that are actually present in your dataset.

Re: monthly data format

SystemsasdD2019 — Tue, 14 Jan 2020 15:27:09 GMT

Sory, But the procedure requires the use of all variables Lsilverprice Lgoldprice and variable of date OUTMONTH .there is no way to combine them , when I apply the program :

title "Illustration of ODS Graphics";
proc varmax data=WANT plot(unpack)=(residual model forecasts);
id OUTMONTH interval=month;
model Lsilverprice Lgoldprice  / dify=(1,12) noint p=0 q=1;
output lead=12;
run;

appears an ERROR: variable Lsilverprice not found

ERROR variable Lgoldprice not found

when I apply the program ;

title "Illustration of ODS Graphics";
proc varmax data=IMPORT plot(unpack)=(residual model forecasts);
id OUTMONTH interval=month;
model Lsilverprice Lgoldprice / dify=(1,12) noint p=0 q=1;
output lead=12;
run;

An error appears:

ERROR : variable OUTMONTH not found

there is no way to combine them , with thanks and appreciation .

Re: monthly data format

Kurt_Bremser — Tue, 14 Jan 2020 15:42:27 GMT

KNOW

YOUR

DATA

You have to know which variables are present in your dataset, and subsequently which ones to use in the procedure, according to the documentation. We could only try to provide further help if you give us a usable example of your dataset in a data step with datalines and what information you want to gain by using the varmax procedure.

Re: monthly data format

Reeza — Tue, 14 Jan 2020 16:29:57 GMT

SAS is saying your input data set, called WANT in your code does not have the variables you're trying to use.

Run a PROC CONTENTS on your data set to verify that you have the variables you need.

proc contents data=want;
run;

If the data set does not have the variables Lgoldprice and Lsilverprice, then there's no data to be used. Check the output in the results pane to see if the variables are present.

You say there's no way to 'combine them'. What's them? What type of combination are you thinking of here?

From the code you've posted earlier your data set only has month so not sure where you're expecting the prices to come from.