I found is easier to read the shape from innermost to outmost, so I won’t misread numbers of “[ ]”. Like in the above, it has 2 columns and 3 row and 3 examples so the shape is (3,3,2)
Second question is what do we call each dimension in the multi-dimension array?
Like above it’s the same example but I added another “[ ]” which makes it to (1,3,3,2)
What do we call this? like in a normal 2d array(3,2), we can read as 3 rows and 2 columns.
The way of reading the shape of an array can be different for each person. Personally, I read the shape from the outermost to the innermost dimensions. However, by simply using .shape you can get the dimensions.
For naming each dimension in a multi-dimensional array, there isn’t a strict standard, but there are common conventions:
1D Array (Vector) → n-vector
2D Array (Matrix) → Dimensions: rows and columns
3D Array (Also called a tensor) → Dimensions: depth, rows, and columns
4D and Higher Dimensional Arrays (Generally referred to as tensors) → Naming conventions can become more complex, with terms like batch, channels, depth, rows, columns, etc., depending on the context of the data.
