矩阵乘法“*”在MATLAB中有效，但在R中无效

I'm trying to multiply a 5 x 3 matrix X by its transpose tX in R.

我试图将一个5 x 3矩阵X乘以它在R中的转置tX。

> X
     [,1] [,2] [,3]
[1,]    1   13  0.5
[2,]    1   23  0.4
[3,]    1    7 -0.2
[4,]    1   16  1.0
[5,]    1   11  0.3

> tX
     [,1] [,2] [,3] [,4] [,5]
[1,]  1.0  1.0  1.0    1  1.0
[2,] 13.0 23.0  7.0   16 11.0
[3,]  0.5  0.4 -0.2    1  0.3

> tX * X
Error in tX * X: array incompatibili

I did this in MATLAB and got immediately the right result. Why can't I do this in R? Thank You.

我在MATLAB中做了这个,并立即得到了正确的结果。为什么我不能在R中这样做?谢谢。

compute this in MATLAB

在MATLAB中计算

1 个解决方案

#1

You need tX %*% X. Or try crossprod(X). See ?"%*%".

你需要tX%*%X。或者尝试crossprod(X)。看?“%*%”。

Actually, if you just read the matrix-multiplication info, you will see this. I added it in two days ago.

实际上,如果你只是阅读矩阵乘法信息,你会看到这一点。我在两天前添加了它。

The "*" is used for Hadamard product, i.e., the element-wise product. Since tX and X don't have the same dimension, you get error. (A hint: I don't know how you get tX. In R, function t transposes a matrix. See transpose info.)

“*”用于Hadamard产品,即元素产品。由于tX和X的维度不同,因此会出现错误。 (一个提示:我不知道你是怎么得到的。在R中,函数t转置一个矩阵。见转置信息。)

I haven't used MATLAB for 10 years. The following docs for R2018a looks different from what I recall.

我没有使用MATLAB 10年。 R2018a的以下文档与我记得的不同。

Anyway, the syntax between two scientific languages are different enough.

无论如何,两种科学语言之间的语法是不同的。

#1