[R] Problem in performing goodness of fit test in R.

(Ted Harding) Ted.Harding at manchester.ac.uk
Sun Feb 14 11:53:39 CET 2010


Good that you got it to work somehow!

Faiz: can you report the result, exactly as returned by R, of

  sum(c(1,1,1,1,1,1)/6) - 1

??

On my Linux with R version 2.10.0 (2009-10-26) this gives 0.
(And, by the way, did you really mean to say you were using
"R 2.1.01"? Was this a typing error for "R 2.10.1"?)

Ted.

On 14-Feb-10 09:26:00, Faiz Rasool wrote:
> Hi  all,
> 
> Following Denis's code, I am able to carry out goodness of fit test.
> 
> However the responses I have received, indicate that the code I
> mentioned in 
> a prior email was not incorrect. inclusion of rep(1,6)/6 made a
> difference 
> for me. I am using windows xp and R 2.1.01. Does difference of
> operating 
> system makes a difference? As without rep(1,6)/6 R would not carry out 
> chisq.test on my computer. The code I got it to work follows.
> 
>> freq=c(22,21,22,27,22,36)#frequencies obtained after rolling the dice
>> 150 
>> times.
>> prob=rep(1,6)/6#change made after Dennis's suggestion.
>> chisq.test(freq,p=prob)#no error message received now.
> 
>         Chi-squared test for given probabilities
> 
> data:  freq
> X-squared = 6.72, df = 5, p-value = 0.2423
> 
> thanks everyone,
> Faiz.
> 
> ----- Original Message ----- 
> From: "Ted Harding" <Ted.Harding at manchester.ac.uk>
> To: <R-help at r-project.org>
> Cc: "Faiz Rasool" <faiz7r at gmail.com>
> Sent: Sunday, February 14, 2010 2:02 PM
> Subject: RE: [R] Problem in performing goodness of fit test in R.
> 
> 
>> On 14-Feb-10 07:42:12, Faiz Rasool wrote:
>>> I am trying to perform goodness of fit test using R. I am using this
>>> website
>>> http://wiener.math.csi.cuny.edu/Statistics/R/simpleR/stat013.html
>>> for help. However, I am unable to carry out the test successfully. My
>>> code follows. It is taken from the website just mentioned.
>>>
>>  freq=c(22,21,22,27,22,36) # frequencies obtained after
>>                            # rolling the dice 150 times.
>>  prob=c(1,1,1,1,1,1)/6 # specify expected frequency for each category.
>>  chisq.test(freq,p=prob) # I do not know what this line means.
>>                          # I just followed instructions on the
>>                          # website.
>>>
>>> The erorr I receive is "erorr in chisq.test(freq,p=prob)/6
>>> probabilities must sum to 1"
>>>
>>> I am very new to R, so any help would be appreciated.
>>> Faiz.
>>
>> I suspect that you must have made an error in entering the commands
>> into R. Prime suspect: You did not have 6 1's in p -- for example
>> you may have put
>>
>>  prob=c(1,1,1,1,1)/6
>>
>> (with only five). I copied your code (as trivially reformatted above)
>> straight into R using copy&paste with the mouse, with results:
>>
>>  freq=c(22,21,22,27,22,36) # frequencies obtained after
>>                            # rolling the dice 150 times.
>>  prob=c(1,1,1,1,1,1)/6 # specify expected frequency for each category.
>>  chisq.test(freq,p=prob) # I do not know what this line means.
>>
>>  #        Chi-squared test for given probabilities
>>  # data:  freq
>>  # X-squared = 6.72, df = 5, p-value = 0.2423
>>                          # I just followed instructions on the
>>                          # website.
>>
>> So it worked as it should work. Therefore something went wrong
>> when you entered the code. Check your input!
>>
>> Hoping this helps,
>> Ted.
>>
>> --------------------------------------------------------------------
>> E-Mail: (Ted Harding) <Ted.Harding at manchester.ac.uk>
>> Fax-to-email: +44 (0)870 094 0861
>> Date: 14-Feb-10                                       Time: 09:02:01
>> ------------------------------ XFMail ------------------------------ 
> 

--------------------------------------------------------------------
E-Mail: (Ted Harding) <Ted.Harding at manchester.ac.uk>
Fax-to-email: +44 (0)870 094 0861
Date: 14-Feb-10                                       Time: 10:53:34
------------------------------ XFMail ------------------------------



More information about the R-help mailing list