drop table agg; create table agg(n integer); begin for i in 0..1000 loop insert into agg values(i); end loop; end; / show errors; select 'Population' d,count(n),avg(n),stddev(n),stddev_pop(n),stddev_samp(n) from agg union select 'Sample',count(n),avg(n),stddev(n),stddev_pop(n),stddev_samp(n) from agg where mod(n,2) = 0 union select 'Sample',count(n),avg(n),stddev(n),stddev_pop(n),stddev_samp(n) from agg where mod(n,10) = 0 union select 'Sample',count(n),avg(n),stddev(n),stddev_pop(n),stddev_samp(n) from agg where mod(n,100) = 0; 得到的結(jié)果是 雖然這些數(shù)據(jù)是人為計(jì)算得到的,并沒有在此錐形曲線上,但是這個(gè)例子反映了_SAMP和_POP的差別。從這個(gè)例子看出,樣本的STDDEV_POP(N)列的數(shù)據(jù)比樣本的STDDE(N)列數(shù)據(jù)更接近實(shí)際的STDDECV(N)列數(shù)據(jù)。