Calculate percentage of positive, zero and negative values

[Agent1592]

Guys:

I have some data for two quarters. I am trying to create a want table like this one. Basically I am trying to get all the possible combinations of positive, negative and zero values for the two quarters. Can someone please help?

Q1	Q2	Percentage of  observations
+	-
+	+
+	0
-	-
-	+
-	0
0	0

data QTR1_QTR2;
  infile datalines dsd truncover;
  input q1:32. q2:32.;
datalines4;
0.0229132564,0.0838041414
0.106891732,0.0074999684
0.0523077125,-0.16913947
-0.038567431,-0.068585141
-0.085454507,-0.243917286
0.0890000313,0.4542936499
0.0110349708,0.083370712
-0.029292079,-0.064858472
0.0099681029,-0.036895711
-0.044827588,0.0155642022
0.0980392184,-0.030651339
0.1329480476,-0.029253897
-0.048689142,-0.044543457
0.0493506626,-0.026244339
0.0932343341,0.0101010271
0.0153109924,0.0172910492
-0.007617369,0.0033984659
-0.01668303,-0.019999972
-0.085653123,0.5458613398
0.1257996012,0.0046728942
-0.098692028,0.0074184272
-0.035149435,-0.073319685
-0.004424775,-0.016517244
0.0091439181,-0.031136883
-0.103247662,-0.003340726
-0.085603985,-0.011174274
-0.01265823,0.1142857201
-0.018691681,-0.021052613
0.0483871303,-0.068747026
-0.03592811,0.0762711838
0.0322581127,-0.03441676
-0.155797125,0.0526315302
-0.019607823,-0.111111035
-0.115241653,0.1209262308
-0.167777808,-0.094501742
0.0526315917,-0.195652182
0.0076775354,-0.013267065
0.0743801304,0.0957983665
-0.003337734,0.0406090978
0.0531119832,-0.044910138
-0.093476823,-0.026143835
0.0443655728,0.0069766844
0.2227414064,0.0838709378
-0.122536445,-0.157512908
0.0226337451,0.0292598979
0.0264391854,-0.093104137
0.0192138422,0.0019499369
-0.015653233,-0.033132504
-0.031857019,0.029294397
-0.002348574,0.0073886328
-0.010393421,0.0396806019
0.0072449318,-0.016953502
-0.025897032,0.0710188485
0.0162396527,-0.006883583
-0.058161403,0.0364925014
0.0279324718,-0.026210079
-0.00850093,0.0102726783
0.0014855785,-0.009168097
-0.013147191,-0.033698457
-0.018088938,0.0199586154
0.0253665383,-0.009128962
-0.008579102,0.0106737093
0.0254237281,0.0243902433
0.0161956251,-0.008353974
0.0301075587,0.0023753507
0.0011792273,0.0131826882
0.0965517395,0
0.2198887777,-0.052272707
-0.006122432,-0.062048489
0.0222310225,-0.099275499
-0.004782655,0.0269315951
0.178947362,-0.111919214
-0.01038965,0.0220640183
0.1296874839,-0.034274949
0.1213826582,-0.001717663
0.0096532162,-0.005925411
-0.010480034,-0.027166404
0.1508723906,-0.097846328
-0.007121461,0.0524017808
0.147407904,0.1645696171
0.0870989747,0.0188849336
-0.034640079,0.0290820163
0.0514352129,-0.05687266
-0.001493591,-0.005280889
-0.046075432,0.0207175256
0.0095069016,-0.055608193
0.0189642295,0.0388938768
-0.02785325,0.0141896625
-0.029350121,0.0227676751
-0.038080272,-0.026218913
-0.033831051,0.0337481215
0.0101404143,0.0059002192
-0.005680872,0.0048869832
0.0207583727,0.0062007108
0.0315314941,0.0019130049
0.0392450775,0.0227726593
-0.001592712,0.0225706998
0.0154698259,0.0430857292
0.0165745868,-0.012903227
-0.056369784,-0.010322571
;;;;

[Astounding · Accepted solution]

First, create a format to perform the grouping:

proc format;

value neg0pos  low-<0 = '-'  0='0' 0-high='+';

run;

If there is the possibility of missing values in your data, you might want to account for those as well.

Then apply the format.  PROC FREQ does a good job counting:

proc freq data=have;

tables q1 * q2 / list missing;

format q1 q2 neg0pos.;

run;

[Agent1592]

Thanks