Compare empirical CDF and lognormal CDF

I have a set of data that follows lognormal distribution. I'm trying to compare lognormal and empirical complementary CDF. However, the lognormal CCDF looks like a horizontal line (please see the attached).

here is the code:

[f_sc ,x_sc]=ecdf(SD); % empirical CDF
fit_s=logncdf(sort(SD) , mean(SD) , std(SD) ); % Lognormal CDF
plot(x_sc,1-f_sc,'-k','LineWidth',3); hold on; 
plot( sort(SD) , 1-fit_s , '-r' , 'LineWidth', 3 ); hold on; 
legend('Empirical CCDF','Lognormal CCDF','FontSize',14,'fontweight','bold','FontName','Time New Roman');

[Ameer Hamza]

What value of SD did you use?

[Mos_bad]

where do you mean? plot or logncdf? I almost tried both logarithmic and actual data values but didn't work. more specific please.

[Ameer Hamza]

I was referring to the value of variable SD in your code. Did you generated it randomly, or do you have a dataset?

Answer 1

[Jeff Miller]

logncdf wants the mean and sd of the logged scores. Try this:

fit_s=logncdf(sort(SD) , mean(log(SD)) , std(log(SD)) );