概率矩阵？P=X/T, T=x[1][1]+x[1][2]+......+x[m][n]?

[此贴子已经被作者于2007-1-26 20:04:52编辑过]

***one way is to recode your 定量数据 into groups and create two new variables called 'weight group' and 'BP group', and do proc freq for these two new variables;

***Another way is to use format to group 定量数据 into different categories. B this way, you don't need to create two new variables***;

data a;

input weight bloodpressure;

cards;

;

run;

proc format;

value wtfmt 0-150='Low' 151-high='high';

value bpfmt 0-120='Normal' 121-high='High';

run;

proc freq data=a;

tables weight*bloodpressure/nopercent nocol chisq;

format weight wtfmt. bloodpressure bpfmt.;

run;