源数据格式(csv):

PM,株系号,重复号,总穗数,有效穗数,实粒数,千粒重,单株产量 g,平均穗长,结实率%,结实率arcsin,GeneType
0,21,c,17,17,1348,25.08,33.81,20,75.35,1.051241064,11
0,21,c,23,23,2046,25.13,51.42,20.1,75.28,1.050389716,11
-,2,a,7,7,971,12.79,12.42,16.6,88.84,1.230151548,1
-,2,a,13,13,1142,22.95,26.21,18.6,84.28,1.163118036,1
+,1,a,8,8,444,17.16,7.62,14.8,53.82,0.823617188,1
+,1,a,12,12,1178,22.89,26.96,17.8,81.64,1.127921808,1

我已经按单株(每行)作了方差分析看单株差异, 现在想用按 株系号+重复号(如1a, 2a, 21c) 取平均值后,针对平均值做方差分析看小区的差异.

应该如何用SAS语句实现?   我也知道用Excel的averageif算好后导出也可以, 但我想学学用SAS实现的方法.