please help me on this ----- Message Text ----- Dear all R users first, sorry for that this question might not be appropriate to ask here. I wanna know theories or techinques aimed at following questions: I have a sample, say,K(at the range from 0 to 20000); the sample data's central moments m(1)---m(j) are estimated(j can be large). also, I can use some methodology to calculate the upper and lower bound of the probabilty of any interested interval, say, for the interval (400--800) with all these information, I wanna recover the distribution of the data, at least recover to some approximating analytic form.Does anybady know such theory or techiniques? your help will be highly appreciated. best regards yong
Spencer Graves
2004-Jul-11 16:04 UTC
[R] Distribution of Data (was: your reference on this problem highly appreciated)
There are many tools for this, e.g., qqnorm, density, and in library(MASS) fitdistr. Also do a literature search on transformations (especially to transformations to normality) and on mixture distributions, esp. Titterington, Smith and Makov (1986) Statistical Analysis of Finite Mixture Distributions (Wiley). What is the nature of your application? If you tell us more about the context, many people could tell you which distributions might be plausible and which would not be credible except as an approximation, e.g., a normal distribution for numbers that can not be negative and whose distribution might be positively skewed. hope this helps. spencer graves p.s. PLEASE do read the posting guide! R-project.org/posting-guide.html Yong Wang wrote:>please help me on this >----- Message Text ----- >Dear all R users >first, sorry for that this question might not be appropriate to ask here. > >I wanna know theories or techinques aimed at following questions: > >I have a sample, say,K(at the range from 0 to 20000); the sample data's >central moments m(1)---m(j) are estimated(j can be large). >also, I can use some methodology to calculate the upper and lower bound of >the probabilty of any interested interval, say, for the interval >(400--800) > >with all these information, I wanna recover the distribution of the data, >at least recover to some approximating analytic form.Does anybady know >such theory or techiniques? > >your help will be highly appreciated. >best regards >yong > >______________________________________________ >R-help at stat.math.ethz.ch mailing list >stat.math.ethz.ch/mailman/listinfo/r-help >PLEASE do read the posting guide! R-project.org/posting-guide.html > >