Re: [Numpy-discussion] Sampling from the multivariate normal

2011-11-09 Thread Joshua Anthony Reyes
Hi Oliver, Robert, Thanks to both of you for your speedy and enlightening responses. They really cleared things up. Now my simulations seem to work as intended. All the best, Josh On 11/9/11 12:14 PM, Robert Kern wrote: > On Wed, Nov 9, 2011 at 16:40, Joshua Anthony Reyes > wrote: >> Hi List

Re: [Numpy-discussion] Sampling from the multivariate normal

2011-11-09 Thread Robert Kern
On Wed, Nov 9, 2011 at 16:40, Joshua Anthony Reyes wrote: > Hi List, > > I'm new to Numpy and I'm a little confused about the behavior of > numpy.random.multivariate_normal(). I'm not sure I'm passing the > variances correctly. My goal is to sample from a bivariate normal, but > the kooky behavior

Re: [Numpy-discussion] Sampling from the multivariate normal

2011-11-09 Thread Olivier Delalleau
The "normal" function takes as input the standard deviation, while "multivariate_normal" takes as input the covariance. -=- Olivier 2011/11/9 Joshua Anthony Reyes > Hi List, > > I'm new to Numpy and I'm a little confused about the behavior of > numpy.random.multivariate_normal(). I'm not sure I

[Numpy-discussion] Sampling from the multivariate normal

2011-11-09 Thread Joshua Anthony Reyes
Hi List, I'm new to Numpy and I'm a little confused about the behavior of numpy.random.multivariate_normal(). I'm not sure I'm passing the variances correctly. My goal is to sample from a bivariate normal, but the kooky behavior shows up when I sample from a univariate distribution. In short,