Data Manipulation with Base SAS

[Singham20]

Hi SAS Community!

I was hoping someone could help me with the following issue.

I have 2 data sets, Dataset 1 and Dataset 2.  I need to use Dataset 2 to manipulate the data in Dataset 1 to create Dataset 3.  Dataset 2 has the Start time of the data and I need all the data that is populated before the Start time for each variable to be changed to blank.  Below is the visuals for this process.

Dataset 1
Time	T1	T2	T3	T4	T5
1/1/2017 0:00	6.818272	4.030038	8.213272	0.060326	9.521269
1/1/2017 1:00	0.100979	3.009709	2.359012	1.749067	8.0581
1/1/2017 2:00	1.20293	0.524906	0.686842	2.561308	3.58268
1/1/2017 3:00	0.178333	8.815986	7.657224	5.581614	1.142414
1/1/2017 4:00	6.654463	2.398171	2.696395	2.181824	2.000096
1/1/2017 5:00	5.428405	7.665879	7.996566	9.112351	3.012801
1/1/2017 6:00	7.710257	1.296825	7.965649	1.278932	8.198926
1/1/2017 7:00	5.733929	8.763634	7.37632	4.684965	6.252313
1/1/2017 8:00	0.173659	8.191108	7.592309	0.764799	1.81264
1/1/2017 9:00	2.782826	3.713065	8.526002	1.06912	9.897541
1/1/2017 10:00	4.46292	4.154905	4.766856	2.918054	1.398898
1/1/2017 11:00	1.949195	1.306601	4.989425	6.146979	6.172791
1/1/2017 12:00	9.865355	5.047937	2.965091	8.109938	0.024781
1/1/2017 13:00	6.361114	0.850972	3.595049	1.18486	3.629694
1/1/2017 14:00	7.494559	7.221369	9.410831	1.467266	2.488785
1/1/2017 15:00	2.129789	5.699575	0.627782	5.550582	0.275467
1/1/2017 16:00	9.046102	5.785753	0.764385	5.695247	6.812834
1/1/2017 17:00	8.056089	2.70044	0.916423	8.568327	3.729957
1/1/2017 18:00	9.318779	7.345054	5.430452	0.23345	0.561244
1/1/2017 19:00	4.827045	9.602832	1.841213	9.62645	8.05601
1/1/2017 20:00	1.623731	9.719941	4.036305	2.812807	7.412819
1/1/2017 21:00	2.199275	7.198554	0.28249	0.405406	6.038437
1/1/2017 22:00	8.08961	6.457083	8.185543	7.05954	5.085864
1/1/2017 23:00	3.040053	7.585161	8.839786	2.42752	7.113683

Dataset 2
Name	Start
T1	1/1/2017 7:00
T2	1/1/2017 2:00
T3	1/1/2017 15:00
T4	1/1/2017 0:00
T5	1/1/2017 20:00

This is the desired output:

Final Dataset
Time	T1	T2	T3	T4	T5
1/1/2017 0:00				0.060326
1/1/2017 1:00				1.749067
1/1/2017 2:00		0.524906		2.561308
1/1/2017 3:00		8.815986		5.581614
1/1/2017 4:00		2.398171		2.181824
1/1/2017 5:00		7.665879		9.112351
1/1/2017 6:00		1.296825		1.278932
1/1/2017 7:00	5.733929	8.763634		4.684965
1/1/2017 8:00	0.173659	8.191108		0.764799
1/1/2017 9:00	2.782826	3.713065		1.06912
1/1/2017 10:00	4.46292	4.154905		2.918054
1/1/2017 11:00	1.949195	1.306601		6.146979
1/1/2017 12:00	9.865355	5.047937		8.109938
1/1/2017 13:00	6.361114	0.850972		1.18486
1/1/2017 14:00	7.494559	7.221369		1.467266
1/1/2017 15:00	2.129789	5.699575	0.627782	5.550582
1/1/2017 16:00	9.046102	5.785753	0.764385	5.695247
1/1/2017 17:00	8.056089	2.70044	0.916423	8.568327
1/1/2017 18:00	9.318779	7.345054	5.430452	0.23345
1/1/2017 19:00	4.827045	9.602832	1.841213	9.62645
1/1/2017 20:00	1.623731	9.719941	4.036305	2.812807	7.412819
1/1/2017 21:00	2.199275	7.198554	0.28249	0.405406	6.038437
1/1/2017 22:00	8.08961	6.457083	8.185543	7.05954	5.085864
1/1/2017 23:00	3.040053	7.585161	8.839786	2.42752	7.113683

Hopefully someone out there can help on this!

Thank you!

[PeterClemmensen · Accepted solution]

Here is one way. A nice problem to use the array lookup

data dataset1;
input Time:anydtdtm. T1-T5;
format Time datetime20.;
infile datalines dlm=',';
datalines;
1/1/2017 0:00,6.818272,4.030038,8.213272,0.060326,9.521269
1/1/2017 1:00,0.100979,3.009709,2.359012,1.749067,8.0581
1/1/2017 2:00,1.20293,0.524906,0.686842,2.561308,3.58268
1/1/2017 3:00,0.178333,8.815986,7.657224,5.581614,1.142414
1/1/2017 4:00,6.654463,2.398171,2.696395,2.181824,2.000096
1/1/2017 5:00,5.428405,7.665879,7.996566,9.112351,3.012801
1/1/2017 6:00,7.710257,1.296825,7.965649,1.278932,8.198926
1/1/2017 7:00,5.733929,8.763634,7.37632,4.684965,6.252313
1/1/2017 8:00,0.173659,8.191108,7.592309,0.764799,1.81264
1/1/2017 9:00,2.782826,3.713065,8.526002,1.06912,9.897541
1/1/2017 10:00,4.46292,4.154905,4.766856,2.918054,1.398898
1/1/2017 11:00,1.949195,1.306601,4.989425,6.146979,6.172791
1/1/2017 12:00,9.865355,5.047937,2.965091,8.109938,0.024781
1/1/2017 13:00,6.361114,0.850972,3.595049,1.18486,3.629694
1/1/2017 14:00,7.494559,7.221369,9.410831,1.467266,2.488785
1/1/2017 15:00,2.129789,5.699575,0.627782,5.550582,0.275467
1/1/2017 16:00,9.046102,5.785753,0.764385,5.695247,6.812834
1/1/2017 17:00,8.056089,2.70044,0.916423,8.568327,3.729957
1/1/2017 18:00,9.318779,7.345054,5.430452,0.23345,0.561244
1/1/2017 19:00,4.827045,9.602832,1.841213,9.62645,8.05601
1/1/2017 20:00,1.623731,9.719941,4.036305,2.812807,7.412819
1/1/2017 21:00,2.199275,7.198554,0.28249,0.405406,6.038437
1/1/2017 22:00,8.08961,6.457083,8.185543,7.05954,5.085864
1/1/2017 23:00,3.040053,7.585161,8.839786,2.42752,7.113683
;

data dataset2;
input Name $ Time:anydtdtm.;
infile datalines dlm=',';
format Time datetime20.;
datalines;
T1,1/1/2017 7:00
T2,1/1/2017 2:00
T3,1/1/2017 15:00
T4,1/1/2017 0:00
T5,1/1/2017 20:00
;

data want(drop= i j name);
   array lookup{5} _temporary_;

   do i=1 by 1 until(eof1); 
      set dataset2 end=eof1;
      lookup[i]=Time;
   end;
   
   array T{5};
   do until(eof2);
      set dataset1 end=eof2;
      do j=1 to dim(lookup);
         if Time lt lookup[j] then T[j]=.;
      end;
      output;
   end;
run;

[Singham20]

Thank you for all your help!  Quick question.  The dataset I showed only had 5 observations.  My actual dataset has thousands each with unique names that is not as simple as T1, T2, and so on.  What can I change the array to so it handles this?

[PeterClemmensen]

@Singham20 can you post some data that represents this? Not with thousands of unique variables though.

[Singham20]

So here is how the variable look. I am just showing 5, however there are 24,785 variables.

Time	_48164_1	_W468s_1	_di4ek452_5	_47816_1	_zidle_45
1/1/2017 0:00	6.818272	4.030038	8.213272	0.060326	9.521269
1/1/2017 1:00	0.100979	3.009709	2.359012	1.749067	8.0581
1/1/2017 2:00	1.20293	0.524906	0.686842	2.561308	3.58268
1/1/2017 3:00	0.178333	8.815986	7.657224	5.581614	1.142414
1/1/2017 4:00	6.654463	2.398171	2.696395	2.181824	2.000096
1/1/2017 5:00	5.428405	7.665879	7.996566	9.112351	3.012801
1/1/2017 6:00	7.710257	1.296825	7.965649	1.278932	8.198926
1/1/2017 7:00	5.733929	8.763634	7.37632	4.684965	6.252313
1/1/2017 8:00	0.173659	8.191108	7.592309	0.764799	1.81264
1/1/2017 9:00	2.782826	3.713065	8.526002	1.06912	9.897541
1/1/2017 10:00	4.46292	4.154905	4.766856	2.918054	1.398898
1/1/2017 11:00	1.949195	1.306601	4.989425	6.146979	6.172791
1/1/2017 12:00	9.865355	5.047937	2.965091	8.109938	0.024781
1/1/2017 13:00	6.361114	0.850972	3.595049	1.18486	3.629694
1/1/2017 14:00	7.494559	7.221369	9.410831	1.467266	2.488785
1/1/2017 15:00	2.129789	5.699575	0.627782	5.550582	0.275467
1/1/2017 16:00	9.046102	5.785753	0.764385	5.695247	6.812834
1/1/2017 17:00	8.056089	2.70044	0.916423	8.568327	3.729957
1/1/2017 18:00	9.318779	7.345054	5.430452	0.23345	0.561244
1/1/2017 19:00	4.827045	9.602832	1.841213	9.62645	8.05601
1/1/2017 20:00	1.623731	9.719941	4.036305	2.812807	7.412819
1/1/2017 21:00	2.199275	7.198554	0.28249	0.405406	6.038437
1/1/2017 22:00	8.08961	6.457083	8.185543	7.05954	5.085864
1/1/2017 23:00	3.040053	7.585161	8.839786	2.42752	7.113683
1/1/2017 23:00	3.040053	7.585161	8.839786	2.42752	7.113683

[PeterClemmensen]

If your data is representative and all of the variables of interest have prefix _, then simply do

data want(drop= i j name);
   array lookup{5} _temporary_;

   do i=1 by 1 until(eof1); 
      set dataset2 end=eof1;
      lookup[i]=Time;
   end;
   
   do until(eof2);
      set dataset1 end=eof2;
      array T{*} _:;
      do j=1 to dim(lookup);
         if Time lt lookup[j] then T[j]=.;
      end;
      output;
   end;
run;