[R] Odp: how to generate data set with different length and calculate the mean?

[aegea]

Petr, 
Thanks for your suggestions. It makes sense, since I don't know how to make
a matrix with different length of rows. 
I have a concern for this problem. I actually deal with a much bigger
dataset e.g. 1000, and each dataset needs to change the number of data in it
according a vector which has 1000 corresponding  different values. It will
be hard to deal with data one by one. Is there a way I can do them together?
Sorry for not making it clear. 

I am thinking I have to use 'for loop' to get a list of vectors. But I am
not sure how to do it efficiently? Thanks again.




 
Petr Pikal wrote:
> 
> Hi
> 
> I have no idea how you could do what you want. I only recommend you to use 
> list instead of matrix as list can incorporate objects with various size
> 
> I am not sure if this is the most elegant way but you can make your matrix 
> a data frame
> 
> ddd<- as.data.frame(data)
> and than use thist
> 
> lapply(ddd, function(x) unlist(list(x)))
> 
> To get list of vectors
> 
> Regards
> Petr
> 
> r-help-bounces at r-project.org napsal dne 01.02.2010 03:46:34:
> 
>> 
>> Hello,
>> 
>> This may be a rare question. I am struggling to solve it. I really
>> appreciate any help or suggestions. Thanks a lot in advance!
>> 
>> 
>> I put my questions between the code to make it clear. The problem I have 
> is:
>> I generated 10 data sets with 8 data for each set. Now I want to change 
> the
>> number of data in each dataset according to a vector 'size' (as 
> follows),
>> that is, each new dataset contains different number of data. How can I 
> do
>> it? After generating the new datasets, how can I seperate the data from 
> two
>> distributions and calculate the sample mean? Thanks a lot. 
>> 
>> 
>> 
>> # generate 10 data sets, each data sets include 8 sample. 4 from N(0, 1) 
> and
>> 4 from N(5, 1)
>> data<- matrix(0,10,8)
>>  th    <- c(0, 5, 1)
>> for(i in 1:10){
>>  data[i,] <- rnorm(8,mean= rep(th[1:2],8/2),sd=th[3])
>> }
>> 
>> # change the number of samples for each data set.  e.g. the first 
> dataset
>> needs to increase to 20, the #first 8 keep the same, add another 12 
> sample
>> (6 from N(0,1) and the other 6 from N(5, 1) ), the second #dataset needs 
> to
>> increase to 10, keep the first 8 the same, generate another 2 (one from
>> N(0,1) and the #other one from N(5,1)),  the third data set does not 
> need to
>> change. etc. 
>> 
>> size=c(20, 10, 8, 14, 16, 12, 8, 80)
>> 
>> 
>> # Since each data set changes to different size, and add different 
> number of
>> data,  for each dataset how #can I calculate the difference of the 
> sample
>> mean from N(0,1) and the sample mean from 
>> #N(5,1) and the pooled standard deviation of two samples. Two 
> difficulties:
>> each new dataset includes #different number of data; another difficulty,
>> when I generated data, the two successive data are 
>> #from different normal distribution, how can I seperate them and 
> calculate
>> the average for each sample #and pooled standard deviation?
>> 
>> 
>> 
>> -- 
>> View this message in context: 
> http://n4.nabble.com/how-to-generate-data-set-
>> with-different-length-and-calculate-the-mean-tp1458420p1458420.html
>> Sent from the R help mailing list archive at Nabble.com.
>> 
>> ______________________________________________
>> R-help at r-project.org mailing list
>> https://stat.ethz.ch/mailman/listinfo/r-help
>> PLEASE do read the posting guide 
> http://www.R-project.org/posting-guide.html
>> and provide commented, minimal, self-contained, reproducible code.
> 
> ______________________________________________
> R-help at r-project.org mailing list
> https://stat.ethz.ch/mailman/listinfo/r-help
> PLEASE do read the posting guide
> http://www.R-project.org/posting-guide.html
> and provide commented, minimal, self-contained, reproducible code.
> 
> 

-- 
View this message in context: http://n4.nabble.com/within-a-matrix-how-to-add-each-column-with-different-length-of-data-tp1458420p1458870.html
Sent from the R help mailing list archive at Nabble.com.