Lognormal fitting with 10-base
显示 更早的评论
I have the statistical toolbox and used it to make a lognormal fitting with e-base. Now I need a lognormal fitting with 10-base, does anyone know how to do this?
采纳的回答
Torsten
2015-3-19
0 个投票
If you are still interested in an answer to your question:
(mu)^= 1/n * sum_{i=1}^{n} Log10[x_i]
and
(sigma^2)^ = Ln[10]/n * sum_{i=1}^{n} (Log10[x_i]-(mu)^)^2
are maximum likelihood estimates for the parameters of your distribution with density function
f(x)=1/(Sqrt[2*Pi*Ln[10]]*Sigma*x)*10^(-(Log10[x]-mu)^2/(2*sigma^2))
Note the normalization by a factor of Sqrt[Ln[10]] to make
integral_{x=0}^{x=Inf} f(x) dx = 1.
Best wishes
Torsten.
更多回答（1 个）
类别
在 帮助中心 和 File Exchange 中查找有关 Lognormal Distribution 的更多信息
Translated by 
