|WikiProject Mathematics||(Rated Start-class, Low-importance)|
|WikiProject Statistics||(Rated Start-class, Low-importance)|
|Daily page views|
formula for the median (table on the right)
The formula for the median in the table on the right "median = square root (ω)" cannot be correct. If so, all curves with the same ω in fig. 2 (Cumulative distribution function) should cross at the same point: cdf(median) = 0,5 I am not so familiar with the Nakagami distribution, therefore I don't know the correct formula. Hartmut Voigt --220.127.116.11 (talk) 12:14, 22 February 2011 (UTC)
(I came here to make a similar comment) Yes, the formula for the median cannot be correct, as it would imply nice closed forms for the median of the chi-square distribution, which are not known to exist. (c.f. chi-square distribution) Shabbychef (talk) 23:09, 3 December 2014 (UTC)
There is difficulty with the online reference to work by Laurenson. This depends on browsers being able to deal with the XBM format which, acoording to XBM is not supported by IE or Mozilla. Is there an alternative? Melcombe (talk) 16:32, 22 July 2011 (UTC)
It seems that in some applications, 0<m<0.5 is possible as well. Maybe this should be mentioned? See eg.:
Dumane, V.A.; Shankar, P.M.; , "Use of frequency diversity and Nakagami statistics in ultrasonic tissue characterization," Ultrasonics, Ferroelectrics and Frequency Control, IEEE Transactions on , vol.48, no.4, pp.1139-1146, July 2001 doi: 10.1109/58.935733
The Nakagami-m and -q distributions are only approximately equal (exactly equal in special cases), and the parameter relationship is more involved than claimed here.
Separately, the Nakagami-m includes as special cases the one-sided Gaussian and the Rayleigh distributions.