循环遍历数据帧的连续列

Question

我想循环遍历一个数据帧的列，并将计算结果存入矩阵。

该场景可以与下面的示例数据被复制：

df = data.frame(replicate(10,sample(0:20,10,rep=TRUE))) # the columns to be calculated on 

M1 = as.data.frame(matrix(0, nrow = 10, ncol = 10)) # a matrix to hold the results. 
rownames(M1) = colnames(df) 
colnames(M1) = colnames(df) 

并且显示如下：

> df # Frame with columns of data, X1 to X10 

    X1 X2 X3 X4 X5 X6 X7 X8 X9 X10 
1 1 19 2 6 6 5 0 2 5 10 
2 16 7 14 16 16 18 11 2 18 11 
3 7 6 11 4 4 1 18 11 10 16 
4 20 2 4 20 4 6 10 5 16 7 
5 9 8 16 19 11 2 14 7 13 7 
6 5 16 6 8 20 15 5 11 4 0 
7 11 16 12 8 18 20 20 20 10 14 
8 17 14 10 4 3 10 13 11 5 1 
9 9 20 10 5 1 7 12 10 5 6 
10 8 14 3 14 20 10 17 20 9 14 

> M1 # Output frame to hold results 

    X1 X2 X3 X4 X5 X6 X7 X8 X9 X10 
X1 0 0 0 0 0 0 0 0 0 0 
X2 0 0 0 0 0 0 0 0 0 0 
X3 0 0 0 0 0 0 0 0 0 0 
X4 0 0 0 0 0 0 0 0 0 0 
X5 0 0 0 0 0 0 0 0 0 0 
X6 0 0 0 0 0 0 0 0 0 0 
X7 0 0 0 0 0 0 0 0 0 0 
X8 0 0 0 0 0 0 0 0 0 0 
X9 0 0 0 0 0 0 0 0 0 0 
X10 0 0 0 0 0 0 0 0 0 0 

在df列X1和X2是在投入到计算，然后X1和X3，然后X1和X4等，然后循环将循环X2和X3，然后循环X2和X4等。

列n和m被输入到计算/循环中，并且结果应该被放置在对应于列的矩阵中的适当位置上，其中 x m。计算本身简单地将Xn和Xm之间的区域确定为绘制线。我不知道如何正确地构建循环来做到这一点：

# The first iteration of the calculation, column X1 and X2 (X1 and X1 would = 0) 

    y = seq(1,10,1) 
    f1 = approxfun(y, df[,1] - df[,2]) # takes two columns as inputs 
    f2 = function(x) abs(f1(x)) 

    area1 = integrate(f2, 1, 10, subdivisions = 500) 
    M1[2,1] = area1$value 

结果框架会产生一个“半矩阵”（即所有需要的镜像一半是相同的）：

> M1 
    X1 X2 X3 X4 X5 X6 X7 X8 X9 X10 
X1 0 0 0 0 0 0 0 0 0 0 
X2 A 0 0 0 0 0 0 0 0 0 
X3 A A 0 0 0 0 0 0 0 0 
X4 A A A 0 0 0 0 0 0 0 
X5 A A A A 0 0 0 0 0 0 
X6 A A A A A 0 0 0 0 0 
X7 A A A A A A 0 0 0 0 
X8 A A A A A A A 0 0 0 
X9 A A A A A A A A 0 0 
X10 A A A A A A A A A 0 

我开始构建一个for循环，但我使用i和j保持在X1，直到它已通过X2-X10循环，然后移动到X2等

感谢绊倒了！

Answer 1

我无法让您的功能正常工作。因此，与使用随机替换功能，这个循环对我的作品：

area=list() # because the actual function doesn't work 
for(i in 1:ncol(df)){ 
    for(j in 1:ncol(df)){ 
    if(i==j){M[i,i]=0;next} 
    selection=df[,c(i,j)] 
    #area=integrate(f2, 1, 200, subdivisions = 500) 
    area$value=mean(colSums(selection)) # something random to check 
    M[i,j]=area$value 
    M[j,i]=area$value 
    } 
} 

但循环一般不做事的最有效的方式。因此，您可能更喜欢此选项：

df = data.frame(replicate(10,sample(0:20,10,rep=TRUE))) # the columns to be calculated on 
my.f = function(x) abs(x[,1]-x[,2]) 

#y = t(as.matrix(combn(ncol(df), 2L, function(y) integrate(my.f(df[y]), 1, 200, subdivisions = 500),simplify=F))) # This doesn't work, but should be close to what you want to do 

y = t(as.matrix(combn(ncol(df), 2L, function(y) mean(f(df[y]),simplify=F)))) # this works, but is just an example 

N = seq_len(ncol(y)) 
nams = colnames(df) 
M = matrix(ncol = length(nams), nrow = length(nams)) 
M[lower.tri(M)] = y 
M = t(out) 
M[lower.tri(M)] = y 
M = t(M) 
diag(M) = 0 
rownames(M) = colnames(out) = colnames(df) 
M 

    X1 X2 X3 X4 X5 X6 X7 X8 X9 X10 
X1 0.0 8.6 6.4 8.8 7.1 6.6 7.0 4.0 7.0 3.7 
X2 8.6 0.0 5.0 4.4 5.5 5.4 4.4 9.2 8.0 7.7 
X3 6.4 5.0 0.0 7.2 5.9 5.8 7.6 7.0 10.4 6.5 
X4 8.8 4.4 7.2 0.0 5.9 4.4 5.4 9.6 8.4 7.3 
X5 7.1 5.5 5.9 5.9 0.0 7.3 5.3 9.1 8.5 8.0 
X6 6.6 5.4 5.8 4.4 7.3 0.0 6.0 8.4 5.6 3.7 
X7 7.0 4.4 7.6 5.4 5.3 6.0 0.0 8.8 4.4 5.7 
X8 4.0 9.2 7.0 9.6 9.1 8.4 8.8 0.0 9.6 6.9 
X9 7.0 8.0 10.4 8.4 8.5 5.6 4.4 9.6 0.0 5.5 
X10 3.7 7.7 6.5 7.3 8.0 3.7 5.7 6.9 5.5 0.0