numpy/matlab compatibility

[Robert Kern]

On 1/25/11 2:13 PM, Matt Funk wrote:
> Hi,
>
> i am fairly new to python. I was wondering of the following is do-able
> in python:
>
> 1) a = rand(10,1)
> 2) Y = a
> 3) mask = Y>  100;
> 4) Y(mask) = 100;
> 5) a = a+Y
>
> Basically i am getting stuck on line 4). I was wondering if it is
> possible or not with python?
> (The above is working matlab code)

You will want to ask numpy questions on the numpy-discussion mailing list 
instead of here.

   http://www.scipy.org/Mailing_Lists

When asking how to replicate a particular MATLAB behavior, please describe (and 
preferably also *show*) what the given MATLAB code does. We're not all familiar 
with MATLAB, and even if we are, we may not know which specific aspect of the 
code you want us to replicate.

You will also want to check out this page:

   http://www.scipy.org/NumPy_for_Matlab_Users

To answer your specific questions:

1) a = numpy.random.random_sample([10, 1])
2) Y = a.copy()
   # In Python assignment to a new name does not copy the object. You seem to 
want a copy in this case. numpy arrays have a .copy() method, though most Python 
objects don't.
3) mask = Y > 0.5
   # Note that Y only has values in [0..1) at this point, so using 100 here will 
create a boolean mask with only False entries. This is probably not what you 
wanted to show, so I used 0.5 instead.
4) Y[mask] = 100
5) a = a + Y
   # Or "a += Y" if you want to modify "a" in-place.

-- 
Robert Kern

"I have come to believe that the whole world is an enigma, a harmless enigma
  that is made terrible by our own mad attempt to interpret it as though it had
  an underlying truth."
   -- Umberto Eco