Re: WeibullDistribution
- To: mathgroup at smc.vnet.net
- Subject: [mg42571] Re: WeibullDistribution
- From: "Robert Nowak" <robert.nowak at ims.co.at>
- Date: Wed, 16 Jul 2003 09:13:33 -0400 (EDT)
- References: <bf093n$v3$1@smc.vnet.net>
- Sender: owner-wri-mathgroup at wolfram.com
Hello Bill, thank you for the nice lesson. now i got it. what about doing a NonlinearFit on the empirical CDF of the data against the CDF of the desired dsitribution. are there any tehoretical issues against such a fit ? of course is see that linear fitting is much more elegant but isn't there a danger to get som bias in the estimated parameters due to the transformations isn't it neccessary to weight the data properly to take the transformations into account ? regards robert "Bill Rowe" <listuser at earthlink.net> wrote in message news:bf093n$v3$1 at smc.vnet.net... > On 7/14/03 at 5:42 AM, robert.nowak at ims.co.at (Robert Nowak) wrote: > > > could you please outline how exactly to do the fit ? based on the array > > called data. > > In[1]:= > <<Statistics` > > In[12]:= > data=RandomArray[WeibullDistribution[5,2],{1000}]; > > In[18]:= > H=Transpose[{Sort@data,-Log[1-(Range[Length at data]-.5)/Length@data]}]; > > In[22]:= > f=Fit[Log@H,{1,t},t]; > {a,b}={f[[2,1]],Exp[-f[[1]]/f[[2,1]]]} > > Out[23]= > {4.910299794620335, 1.9841929305816408} > > Note, in my earlier post there was an error I failed to notice in the code for H. The discription of H as the cumulative hazard function equal to -Log[1-CDF] was correct but the code computed -Log[CDF] which is incorrect. >