[SciPy-dev] NumPy_for_Matlab_Users page drawbacks

[dmitrey]

I think the page still has some problems:

1) repmat entry: python equivalent is tile(), not repmat().
Also check for correct** **numpy.matrix entry for repmat is needed (I 
don't know is all ok there)
2) I think the page misses entry for c = [a b] equivalent for flat 
arrays (btw I need the one for my purposes)
3) everywhere is numerical equivalents, like

a(1:3,5:9) -> a[0:3][:,4:9]

But in size, I don't know why, symbolic eqivalent is implemented:
size(a,n) -> a.shape[n]
# btw I still think there should be shape[n-1], or, better, size(a,2) -> 
a.shape[1]
There is a comment in 4th column about 0 vs 1 indexing, but who really 
reads the column?!

D.