[R] moc function, normal distribution peak capture

[huan.huang@uk.bnpparibas.com]

Hi

With moc function (in moc library), I try to apply a mixture of 2 normal
distribution for univariate data which has 2 peaks,
however I found it came down to 1 normal distribution (weights became 1 and
0) and could not capture the 2nd peaks in many cases.
I set the first mode and second mode as initial values of mean of normal
distribution, and set an appropriate sd as initial values of sd of normal
distribution.

Has anyone experienced this problem? If anyone could let me know how to
avoid this problem, it would be appreciated.

Thanks

Huan


This message and any attachments (the "message") is\ intende...{{dropped}}