Getting plotted x and y points from probplot

I am making a probplot of some data and would like to find the x-value corresponding to a given probability, e.g. prob>0.99 what is the corresponding value on the x axis.

Taking the output of a probplot function call

ax = probplot(data, [], [], 'noref'); one would think the datat would be in ax.XData and ax.YData but this is very misaligned with respect to what is plotted, e.g. y data not going from 0 to 1. Is there a way to extract the transformed data?

Answer 1

[Alan Moses]

ax = probplot(data) returns a graphics array. The ‘x’ and ‘y’ data can be accessed by the following lines:

ax(1).XData %returns the x data
ax(1).YData %returns the y data

The ‘y’ data represents the quantiles of the distribution. By default, the normal distribution is considered. The quantiles are converted into probability values. In case of uncensored data, the probability values can be found using the formula (i-0.5)/N, where N is the number of data points and ‘i’ runs from 1 to N.

You may refer to the “Algorithms” section in the documentation that explains the same. You may also refer to the post here.