problem with lognormal cumulative distribution function

2 次查看（过去 30 天）
Karim Tarbali
Karim Tarbali 2011-9-4
Hi. i want to get the lognormal cumulative distribution function for a variable with mean=0.04 and stdv=0.01 :
cdf('Lognormal',value,mean,stdv)
cdf('Lognormal',0.04,0.04,0.01)
but for the numbers around the mean value the CDF is zero which it should not be. could you please help me what is the prolem

请先登录，再回答此问题。

采纳的回答

bym
bym 2011-9-4
you need to take the exponential of number:
logncdf(exp(.04),.04,.01)
ans =
0.5000
  2 个评论
Karim Tarbali
Karim Tarbali 2011-9-4
thank you so much for you reply. so sorry to bother but why in this syntax the ans=0.5 does not appear:
logncdf(exp(26061000),26061000,0.15*26061000)

请先登录，再进行评论。

更多回答（2 个）

the cyclist
the cyclist 2011-9-4
The support for the density function lies in a narrow band around log(x)-mu = 0, not around x - mu = 0. Here are some values around the peak of the pdf:
mu = 0.04;
sigma = 0.01;
x = [1.01 1.02 1.03 1.04 1.05 1.06 1.07]
cdf__of_lognormal_evaluated_at_x = cdf('Lognormal',x,mu,sigma)
Oleg Komarov
Oleg Komarov 2011-9-4
mu = 0.04;
sigma = 0.01;
values = 0:0.01:2;
subplot(211)
plot(values,pdf('logn',values,mu,sigma))
subplot(212)
plot(values,cdf('logn',values,mu,sigma))
http://en.wikipedia.org/wiki/Log-normal_distribution
As you can see from wikipedia, the smaller the sigma, the more the distribution converges in probability to exp(mu).

类别

在 Help Center 和 File Exchange 中查找有关 Lognormal Distribution 的更多信息

标签

Community Treasure Hunt

Find the treasures in MATLAB Central and discover how the community can help you!

Start Hunting!

Translated by