HELP! check my coding please for histogram with overlay of normal distribution?

x=[23948,25977, 29136, 33815, 36435, 40591, 43243, 44691, 44805, 45509, 44830, 45094, 44236, 43467, 44189, 45846, 48375, 48285, 46372, 44667, 41938]; [n,x] = hist(x,12); %Compute histogram hbar = bar(x, n/trapz(x,n)); %Bar plot of histogram hold on

 %Overlay theoretical Gaussian curve
 Ytheory = normpdf(x,41211.85714,50732773.43);
 htheory = line(x,Ytheory,'Color','r','LineWidth', 2);

(1) Can anyone explain whi i don't get a proper Gaussian curve ontop of my histogram and it just looks like a line?

NOTE:The mean and variance values I used for the normal are the sample mean and variance. If possible check these in case this is the problem (as variance seems VERY large).

(2)Also have I rescaled my histogram (so distribution under curve has area 1) so it can be compared with the distribution correctly?

[José-Luis]

Please edit your post so it is easily readable. Caps are interpreted as shouting in most places.

[José-Luis]

Well, for starters, normpdf expects the third argument to be the standard deviation and you are feeding it the variance of x.

[mabdorab]

Jose-Luis Thank you so much, its fixed it for me.

Suppose I wanted to compare my data to a chi2pdf(x,V) Would I use my mean value as an estimator for V or is V (21-1) because there are 21 different observation years.

[José-Luis]

I don't get it. You could test whether your sample comes for chi-squared distribution. The number of data points would only influence the statistic you are comparing against.