On Oct 27, 2009, at 7:43 AM, Raspaud Martin wrote:

-----BEGIN PGP SIGNED MESSAGE----- Hash: SHA1

Hello,

I’m using numpy v1.2.0, and I have the following codes that provide different results :

---

cal = (PyArrayObject *)PyArray_SimpleNew(2,dims,NPY_FLOAT); for(i=0;i<dims[0];i++) for(j=0;j<dims[1];j++) { *((npy_float *)PyArray_GETPTR2(cal,i,j))=(npy_float)in[i][j]; }

---

and

---

cal = (PyArrayObject *)PyArray_SimpleNewFromData(2,dims,NPY_FLOAT,in);

---

As you probably guessed, "in" is a 2D array of floats of dimensions "dims".

My questions are thus:

• • Why do the two methods provide different results ?
• • How do I get the second to behave like the first ?

In the second case, "in" should be a pointer to a place in memory with space for dims[0]*dims[1] floats. In particular, it should not be a 2-d array of floats.

FromData expects to get a single pointer to float (not a 2D array). I can't think of a way to get the second case to work other than have "in" be a 1-D array.

-Travis