AFEchidna包内置函数之summary

yzhlinscau

已于 2022-05-01 19:10:33 修改

阅读量236

点赞数

分类专栏： R Echidna 文章标签： AFEchidna

于 2022-05-01 17:23:01 首次发布

本文链接：https://blog.csdn.net/yzhlinscau/article/details/124529290

版权

Echidna 同时被 2 个专栏收录

53 篇文章 6 订阅

订阅专栏

50 篇文章 4 订阅

订阅专栏

== Note: AFEchidna包最新版本为1.62，github版本暂未实时更新。==

update()用法如下：

// summary
summary(object)

返回的结果为列表，含有5个元件：
org.res，varcomp，IC，coef.fixed和coef.random。

简单示例如下：

res11<-echidna(h3~1+Rep,
               random=~Fam,
               residual=NULL,
               es0.file="fm.es0")

运用summary()来获取相应元件结果：

// orginal result from .esr files
> cat(summary(res11)$org.res)
 Akaike Information Criterion   4697.68 (assuming 2 parameters).
 Bayesian Information Criterion 4706.31

          Analysis of h3 

                         Wald F statistics
Source of Variation           NumDF     DenDF     F-inc     F-con    P-inc 
 mu                               1      53.4  11253.79  11252.85    <.001
 Rep                              4     525.3     36.34     36.34    <.001

 Model_Term                     Order     Gamma         Sigma     Z_ratio  %C
 Fam                               55  0.834411E-01   132.587        2.35   0 P    
 Residual_units                   559   1.00000       1588.99       15.84

  mv                                       0 effects fitted.
  Fam                                     55 effects fitted.

 Note: Some residuals are large (0 -ve, 4 +ve); 0 are very large.
 Note: Denominator DF calculation may be expensive. Use !DDF -1 to suppress.

其它结果：

// An highlighted block
> summary(res11)$varcomp
      Term   Sigma      SE  Z.ratio
1 Residual 1589.00 100.290 15.84405
2      Fam  132.59  56.459  2.34843
> summary(res11)$IC
   DF parNO     LogL     AIC     BIC
1 554     2 -2346.84 4697.68 4706.31
> summary(res11)$coef.fixed
  Term Level  Effect    SE
1  Rep     1   0.000 0.000
2  Rep     2 -11.732 5.641
3  Rep     3  46.779 5.534
4  Rep     4  23.807 5.521
5  Rep     5   4.895 5.510
6   mu     1 231.494 4.389
> summary(res11)$coef.random %>% head
  Term Level Effect    SE
1  Fam 70048  3.792 7.781
2  Fam 70017 -0.903 8.574
3  Fam 70002  2.552 8.399
4  Fam 70010  3.489 7.786
5  Fam 70041  2.947 8.760
6  Fam 70031  6.637 8.393

当进行性状的批量分析时，summary()可以很方便地得到随机效应和固定效应的结果。

简单示例如下：

> res11b <- update(res11,trait=~h3+h4+h5,batch=T)
> summary(res11b)$coef.fixed
$h3
  Term Level  Effect    SE
1  Rep     1   0.000 0.000
2  Rep     2 -11.732 5.641
3  Rep     3  46.779 5.534
4  Rep     4  23.807 5.521
5  Rep     5   4.895 5.510
6   mu     1 231.494 4.389

$h4
  Term Level  Effect    SE
1  Rep     1   0.000 0.000
2  Rep     2  -5.972 8.549
3  Rep     3  83.994 8.315
4  Rep     4  82.364 8.294
5  Rep     5  -0.468 8.280
6   mu     1 366.740 6.511

$h5
  Term Level  Effect     SE
1  Rep     1   0.000  0.000
2  Rep     2   2.022 10.421
3  Rep     3 109.833 10.138
4  Rep     4  93.284 10.114
5  Rep     5 -11.507 10.095
6   mu     1 547.771  8.038

> summary(res11b)$coef.random %>%
+   purrr::map(head)
$h3
  Term Level Effect    SE
1  Fam 70048  3.792 7.781
2  Fam 70017 -0.903 8.574
3  Fam 70002  2.552 8.399
4  Fam 70010  3.489 7.786
5  Fam 70041  2.947 8.760
6  Fam 70031  6.637 8.393

$h4
  Term Level Effect     SE
1  Fam 70048  8.999 11.096
2  Fam 70017  9.229 12.104
3  Fam 70002  1.856 11.885
4  Fam 70010  0.051 11.102
5  Fam 70041  0.146 12.335
6  Fam 70031  9.497 11.877

$h5
  Term Level Effect     SE
1  Fam 70048 14.502 14.233
2  Fam 70017  8.975 15.680
3  Fam 70002  7.072 15.360
4  Fam 70010 -4.218 14.241
5  Fam 70041 -5.097 16.018
6  Fam 70031  6.479 15.349