Fitting weibull parameters when sample x data that is not perfectly continuous

I have data that I would like to fit to a weibull function. However, I am having trouble properly fitting weibull parameters to this data because I can only input y data using wblfit and the x data are not continuous. First, the x data do not start at zero. Second, while x generally increases, it does not increase by a consistent interval throughout the data set. When I just enter the y data, the predicted parameters are clearly inaccurate. Is there any way to fit weibull parameters with data of this nature?

Answer 1

[Walter Roberson]

As discussed in your previous Question:

Weibull is a probability distribution. Each item is independent of the previous items. There are no coordinate pairs, just randomly generated values.

Therefore it does not make sense to talk about x values for wblfit. If you have a proposed function that depends upon x to determine the random outcome, then whatever it is is not a Weibull distribution, and you cannot use wblfit to analyze it.

Answer 2

[John D'Errico]

You asked the same question before. I think you misunderstand the idea of estimating a Weibull distribution from sample data.

wblfit is NOT a tool to fit a weibull PDF to a curve, as if you were doing a nonlinear regression. If you have (x,y) pairs, that is apparently what you are trying to do. You CANNOT use wblfit for that purpose.