Question

我的数据在R控制台中如下所示：

dim（df1）

                        marker X1 X73 X88 X9 X17 X25 X33 X41 X49
1      1228104|F|0-8:C>T-8:C>T  0   0   0  0   0  NA   0   0   0
2    2277607|F|0-10:A>T-10:A>T NA   0   0 NA  NA  NA  NA   0   0
3  100023173|F|0-47:G>A-47:G>A  0   0   0 NA  NA  NA   0   0   0
4      1043336|F|0-7:A>G-7:A>G  1   1   1  0  NA   0   1   1   1
5    1212218|F|0-49:A>G-49:A>G  0   0   0  0   0   0   0   0   0
6    1019554|F|0-14:T>C-14:T>C  0   0   0  0  NA   0   0   0   0
7    1114675|F|0-18:T>C-18:T>C  0   0   0  0   0   0   0   0   0
8  100024550|F|0-16:G>A-16:G>A NA  NA  NA NA  NA  NA   0   0   0
9    1271969|F|0-22:T>A-22:T>A  0   0  NA  0  NA  NA   0   0   0
10     1106702|F|0-8:C>A-8:C>A  0  NA   0  0   0   0   0   0  NA

df1 [1:10，1:10]

## % of missing per genotypes/samples
pmg <- apply(df1, 2, function(gid) sum(is.na(gid)) / length(gid))
length(pmg)

## dropping bad genotypes/ samples
df2 <-  data.frame(marker=df1[,1], df1[,pmg <= .2][,-1])
dim(df2)
#[1] 54003   909

## % of missing per snp
pms <- apply(df1[pmg <= .2, ], 1,
             function(snp) sum(is.na(snp)) / length(snp))
hist(pms)
length(pms)

# removing bad snps with high missing values 
df3 <- df2[pms <=0.2,]
dim(df3)
# [1] 37982   909

我根据此（FILTER1）过滤数据

## % of missing per snp
pms <- apply(df1[pmg <= .2, ], 1,
             function(snp) sum(is.na(snp)) / length(snp))
hist(pms)
length(pms)

# removing bad snps with high missing values 
df2 <- df1[pms <=0.2,]
dim(df2)


## % of missing per genotypes/samples
pmg <- apply(df1, 2, function(gid) sum(is.na(gid)) / length(gid))
length(pmg)

## dropping bad genotypes/ samples 
df3 <-  data.frame(marker=df2[,1], df2[,pmg <= .2][,-1])
dim(df3)

我尝试运行相同的过滤器，但是这次我按照以下步骤切换了两个步骤的顺序（过滤器2）：

> pms <- apply(df1[pmg <= .2, ], 1,
>              function(snp) sum(is.na(snp)) / length(snp))

当我在FILTER 2中运行此代码时；

[.data.frame

它给了我这个错误：

protractor.Key.COMMAND（df1，pmg <= 0.2，）中出现错误：找不到对象'pmg'

如您所见，我切换了几个步骤，因此也应该修改代码才能正常工作。但是我不知道怎么做。

Answer 1

使用前，您需要定义pmg。只需将本节移到您的代码顶部即可：

## % of missing per genotypes/samples
pmg <- apply(df1, 2, function(gid) sum(is.na(gid)) / length(gid))
length(pmg)

R中丢失数据的百分比给出错误

1 个答案: