Time Series Frequency Transformation (Year to Quarter)

Toni2 — Wed, 23 Jun 2021 16:02:56 GMT

hi, i am trying to understand how the variable X can be converted from annual data to quarterly. (Please note that i have already the X variable in quarters but i need to understand how it has been converted)

i used Xy which is on annual basis and then using the proc expand tried to transform it to quarterly data.

However, the outcome from SAS is not what i expected. At the bottom i have attached the X variable measured in quarters (this is what i expect from SAS)

data a;
input year $4. Xy;
datalines;
2000 519859
2001 561439
2002 606166
2003 656433
2004 713004
2005 770328
2006 836010
2007 903466
2008 951003
2009 971475

;
run;

data b;
set a;
year1 = mdy(1,1,input(year,4.));
format year1 year4.;
drop year;
rename year1=year;
run;

proc expand data=b out=c from=year to=quarter;
id year;
convert xy = xq_tot / observed = total;
convert xy = xq_avg / observed = average;
run;

this is the SAS results. Compare these results with the actual (what i am expecting to see) which are at the bottom there is a difference

Obs	year	xq_tot	xq_avg
1	2000:01:00	125099.3	504716
2	2000:02:00	127895.6	514664
3	2000:03:00	132067.7	524772
4	2000:04:00	134796.4	535062.9
5	2001:01:00	134477.7	545388.6
6	2001:02:00	138591.9	555823.4
7	2001:03:00	142809.7	566562.2
8	2001:04:00	145559.8	577571.8
9	2002:01:00	145118	588691.7
10	2002:02:00	149552.4	599999
11	2002:03:00	154187.9	611706.6
12	2002:04:00	157307.7	623819.9
13	2003:01:00	156997.6	636238.5
14	2003:02:00	161941.6	649124.9
15	2003:03:00	167048.3	662784
16	2003:04:00	170445.5	677066.2
17	2004:01:00	171992.8	691544.5
18	2004:02:00	175439.9	705928.8
19	2004:03:00	180946.6	720114.3
20	2004:04:00	184624.7	734118.2
21	2005:01:00	184247.7	748018.9
22	2005:02:00	190029.7	762219.8
23	2005:03:00	196016.6	777272.5
24	2005:04:00	200034.1	793227.8
25	2006:01:00	199695.3	809730.8
26	2006:02:00	206098.6	826752.3
27	2006:03:00	212817.8	844454.2
28	2006:04:00	217398.3	862430.9
29	2007:01:00	217029.6	879876.1
30	2007:02:00	223564.9	896367
31	2007:03:00	229796.5	911661.8
32	2007:04:00	233075	925369.2
33	2008:01:00	233297.7	937311.2
34	2008:02:00	235622.9	947506.6
35	2008:03:00	240203.1	956061.7
36	2008:04:00	241879.3	962945.7
37	2009:01:00	237977.5	968056.1
38	2009:02:00	241768.3	971449.5
39	2009:03:00	245424.4	973167.1
40	2009:04:00	246304.8	973152.7

Below is the X variable measured in quarters (this is what i am trying to replicate with SAS)

	X
2000 Q1	492537
2000 Q2	500653
2000 Q3	509423
2000 Q4	519859
2001 Q1	529811
2001 Q2	542744
2001 Q3	552252
2001 Q4	561439
2002 Q1	571788
2002 Q2	587243
2002 Q3	593962
2002 Q4	606166
2003 Q1	618291
2003 Q2	631968
2003 Q3	639026
2003 Q4	656433
2004 Q1	667726
2004 Q2	683069
2004 Q3	696401
2004 Q4	713004
2005 Q1	724853
2005 Q2	745101
2005 Q3	756804
2005 Q4	770328
2006 Q1	786507
2006 Q2	805301
2006 Q3	821162
2006 Q4	836010
2007 Q1	852850
2007 Q2	872706
2007 Q3	887064
2007 Q4	903466
2008 Q1	920687
2008 Q2	939411
2008 Q3	941561
2008 Q4	951003
2009 Q1	952468
2009 Q2	960718
2009 Q3	963883
2009 Q4	971475

Re: Time Series Frequency Transformation (Year to Quarter)

Kurt_Bremser — Wed, 23 Jun 2021 13:53:50 GMT

Please post example data as data steps with datalines, so we really know what we are dealing with.

Re: Time Series Frequency Transformation (Year to Quarter)

Toni2 — Wed, 23 Jun 2021 16:03:33 GMT

hi, thanks, i have now edited my post, i hope it helps

topic Time Series Frequency Transformation (Year to Quarter) in New SAS User

Time Series Frequency Transformation (Year to Quarter)

Re: Time Series Frequency Transformation (Year to Quarter)

Re: Time Series Frequency Transformation (Year to Quarter)