chinaathena 发表于 2010-5-2 19:59 
看着书想了很久不知道,网上也找不到相关的资料,脑残了
Here is the program to do std normal, t1 t5 and t30.
Note: t -- distribution with df=1 is cauchy distribution that has very heavy tails( No any order of moments exist)
Hope yuo can do the rest to prove that your are NOT 脑残了.
data t1;
do x=-10 to 10 by 0.1;
pdf_n=pdf('normal', x,0,1);
pdf_t1=pdf('t', x,1);
pdf_t5=pdf('t', x,5);
pdf_t30=pdf('t', x,30);
output;
end;
run;
symbol i=j c=red;
symbol2 i=j c=black;
symbol3 i=j c=green;
symbol4 i=j c=orange;
proc gplot data=t1;
plot (pdf_n pdf_t1 pdf_t5 pdf_t30) *x /overlay;
run;
quit;