[Numpy-discussion] np.linalg.lstsq with several columns all 0 => huge x ?
denis-bz-gg at t-online.de
Thu Oct 25 03:49:40 EDT 2012
On 24/10/2012 19:59, josef.pktd at gmail.com wrote:
> On Wed, Oct 24, 2012 at 1:33 PM, denis<denis-bz-gg at t-online.de> wrote:
>> np.linalg.lstsq of a random-uniform A 50 x 32 with 3 columns all 0
>> returns x[:3] 0 as expected,
>> but 4 columns all 0 => huge x:
> lstsq has rcond argument to do (I think) essentially the same.
> might need to be increased in your example.
rcond 1e-3 works, 1e-6 does not, for these sizes;
the returned singular values don't change much from 3 to 4
columns of 0s, so why the big jump ?
More information about the NumPy-Discussion