Because @zaynah posted in the comments that the data are thought to follow a Weibull distribution, I'm gonna provide a short tutorial on how to estimate the parameters of such a distribution using MLE (Maximum likelihood estimation).
Download and install R, it's free
Optional: Download and inst...
@gung Thanks. I'm not aware that qqPlot from car calculates the MLE parameters automatically. If I generate a random variable with a weibull distribution (rweibull) and use the command qqPlot(rw, distribution="weibull") I get an error message saying that must provide the parameters shape and scale to qqPlot. Am I missing something?
hi, i found that after i import mydata into R, when i do the command,fitdistr(mydata, densfun="weibull") it says error messgae that "mydata" not found.. in fact my data hs been imported into R. any answer would be welcome.
@zaynah Could you please edit your answer and post your code that you use to import the data. Please add the error message too. Could you import the data without errors? Did you check if the data was imported correctly?
i import successfully my data..:read.table("C:/Users/Zaynah/Desktop/mydata.csv"), then i got my data. i mean i saw the full data.. then i use the function > fitdistr(mydata, distribution="weibull") then the error was found:"Error in fitdistr(mydata, distribution = "weibull") : object 'mydata' not found"
@zaynah That's because mydata refers to the whole dataset. What is your data inside the dataset called? If it is called wind.speed, for example, you have to type mydata$wind.speed to access the variable inside your dataset. Just replace wind.speed with your actual name.
@zaynah Please post your R-output in your question (click the edit button below your question). In any case, mydata should have exactly 1 column and n rows (n = your sample size). This column has a name. Type str(mydata) to get it. Use this name then in this way: fitdistr(mydata$myname, distribution="weibull").
read.table("C:/Users/Zaynah/Desktop/mydata.csv") V1 1 5.28,5.28,4.75,3.69,3.69,5.28,5.81,4.22,5.28,5.28,4.22,3.69 2 2.64,4.75,3.69,4.22,3.69,3.69,5.28,5.28,4.75,4.22,4.22,4.22 3 3.17,3.69,3.69,3.69,4.75,4.75,5.28,5.81,4.22,4.75,4.22,4.22 4 4.22,3.69,3.17,4.22,2.64,3.69,4.22,4.75,4.75,4.22,4.22,3.69 5 3.69,3.69,5.28,4.75,4.22,5.23,5.12,5.81,5.81,5.81,4.43,4.43 ps: the edit button is not working...when i do "str(mydata)" Error in str(mydata) : object 'mydata' not found
I think you need to use the option header = TRUE if you have column names (variable names in your csv-file): "mydata <- read.table("C:/Users/Zaynah/Desktop/mydata.csv", header=TRUE)"
And how is your file delimited? By tab? Then you also have to specify that in "read.table" by typing the option "sep="\t""
from research papers i found that they compared the observed data with the Weibull distribution on a graph of probabiliy density distribution against monthly mean speed
The first thing is to correctly format your data file in excel. Please do it like that: The first column contains all wind speeds, from all years and all months. The second column contains the year and the third column contains the months.
I think you see inside the R-Script how I did it. I used fitdistr to calculate the MLE estimate of the shape and scale and then plottet the Weibull density with these parameters (blue curve). Then I've added the density of your data.
Discussion between COOLSerdash and za…
Discussion between COOLSerdash and za…