i'm confused around the screen matrix can i do this with a 2x2 matrix? how do i do it? it's not a 2d transformation matrix if it's 2x2. that would be 3x3. when i think of the 2x2 matrix, i imagine it applying to an equation: the first column multiplies, the second adds. then i can put input things in. maybe it isn't matrix multiplication i'm thinking of. maybe it's just normal broadcast multiplication. or something? i think i would need the points expressed as a 2x2 matrix themselves, for this to work. with 1s underneth them to add. uhhhhhhhhhhhhh what way makes sense to do here? - i have two data items in the form of a vector [x,y,w,1] - i care only about x and y - i want to perform a linear transformation to both x and y - the vectors i have are in a matrix stacked [v1,v2,v3,v4] so what makes sense here? we could do it in two operations, multiply and add there's likely a way to do this with a matrix there are likely other operations that would do multiply+add in numpy, but they may be less intuitive to read in the context of graphics code. so maybe i'll make a 3x3 matrix instead of 2x2. that might make sense. or a 3x2.