Re: [R] How to estimate the parameter for many variable?

2021-07-10 Thread SITI AISYAH ZAKARIA
Dear Rui and Jim, Thank you very much for your feedback. Yes, now I get the output. And after this I will use this output as the marginal distribution to continue the analysis on spatial extremes. Thanks again. See you later. On Fri, 9 Jul 2021 at 17:18, Rui Barradas wrote: > Hello, > > With

Re: [R] How to estimate the parameter for many variable?

2021-07-09 Thread SITI AISYAH ZAKARIA
Dear Rui ang Jim, Thank you very much. Thank you Rui Barradas, I already tried using your coding and I'm grateful I got the answer. ok now, I have some condition on the location parameter which is cyclic condition. So, I will add another 2 variables for the column. and the condition for locati

Re: [R] How to estimate the parameter for many variable?

2021-07-09 Thread Rui Barradas
Hello, With the condition for the location it can be estimated like the following. fit_list2 <- gev_fit_list <- lapply(Ozone_weekly2, gev.fit, ydat = ti, mul = c(2, 3), show = FALSE) mle_params2 <- t(sapply(fit_list2, '[[', 'mle')) # assign column names colnames(mle_params2) <- c("location",

Re: [R] How to estimate the parameter for many variable?

2021-07-08 Thread Rui Barradas
Hello, The following lapply one-liner fits a GEV to each column vector, there is no need for the double for loop. There's also no need to create a data set x. library(ismev) library(mgcv) library(EnvStats) Ozone_weekly2 <- read.table("~/tmp/Ozone_weekly2.txt", header = TRUE) # fit a GEV to

Re: [R] How to estimate the parameter for many variable?

2021-07-08 Thread SITI AISYAH ZAKARIA
Dear all, Thank you very much for the feedback. Sorry for the lack of information about this problem. Here, I explain again. I use this package to run my coding. library(ismev) library(mgcv) library(nlme) The purpose of this is I want to get the value of parameter estimation using MLE by appl

Re: [R] How to estimate the parameter for many variable?

2021-07-08 Thread Jim Lemon
Hi Siti, I think some progress has been made. You have a data set with 888 rows and 19 columns: ow2<-read.table("Ozone_weekly2.txt", header=TRUE) dim(ow2) [1] 888 19 The values may be parts per million ozone in the atmosphere. The columns may represent different measuring locations and my gues

Re: [R] How to estimate the parameter for many variable?

2021-07-08 Thread Rui Barradas
Hello, Also, in the code x <- data.matrix(Ozone_weekly) [...omited...] for(i in 1:nrow(x)) + { for(j in 1:ncol(x)) + {x[i,j] = 1}} not only you rewrite x but the double for loop is equivalent to x[] <- 1 courtesy R's vectorised behavior. (The square parenthesis are needed to keep t

Re: [R] How to estimate the parameter for many variable?

2021-07-08 Thread Jim Lemon
Hi Siti, I think we need a bit more information to respond helpfully. I have no idea what "Ozone_weekly2" is and Google is also ignorant. "gev.fit" is also unknown. The name suggests that it is the output of some regression or similar. What function produced it, and from what library? "ti" is known

[R] How to estimate the parameter for many variable?

2021-07-07 Thread SITI AISYAH ZAKARIA
Dear all, Can I ask something about programming in marginal distribution for spatial extreme? I really stuck on my coding to obtain the parameter estimation for univariate or marginal distribution for new model in spatial extreme. I want to run my data in order to get the parameter estimation val