T O P I C R E V I E W |
b.harris |
Posted - 10/10/2005 : 12:02:40 AM Origin Version (Select Help-->About Origin): 7.5 Operating System: XP
I am needing a little clarification regarding the meaning of the parameters for the lognormal equation.
From the sample curve, it appears that xc is the mode, w the width, and A the value of the mode above y0 (the baseline). However after running this equation several times, and looking at the formula from other sources I gather this is not quite right?
By my reckoning: y0 = offset of baseline w = sqrt of variance (Standard Dev)- a measure of width xc = median (not the mode or mean) and A = The area under the curve if y0 = zero (which can be demonstrated by integrating the curve using the analysis -> calculus -> integrate)
Is this correct??
Many thanks
Ben |
2 L A T E S T R E P L I E S (Newest First) |
kvidovic |
Posted - 06/08/2022 : 02:55:40 AM Hi,
I am also struggling with these parameters. So if I am understood it correctly, xc splits the distribution in two half's. Than xc represents the median or the geometric mean?
If not can you please explain how to calculate the median or geometric mean from xc.
Thank you.
k.v |
Leo_Li |
Posted - 10/10/2005 : 01:44:21 AM Hello b.harris,
You're right. The expression of Origin's Log-normal function is fine but the meanings of xc and w are mis-leading.
i) xc, this parameter doesn't correspond to the Y maximum. The reason is that Log-normal is asymmetric peak function. xc corresponds to the peak center when the Log-normal is approximately symmetric (when w is small). In another word, xc is the "distorted" peak center.
ii) w, this is an asymmetric factor. The peak will be approximately symmetric when w is small.
iii) In some literature, xc and w are also called "mean" and 'standard deviation" respectively. (ref: http://en.wikipedia.org/wiki/Log-normal_distribution )
Leo
|
|
|