How do I fit a gaussian to a bar plot and extract the variance of the gaussian?

Because I'm not starting with raw histogram data, it's my understanding that histfit isn't apropriate for generating a gaussian fit. This has been my approach thus far:

x=[8,11,8,10,8,21,26,33,28,8,17,7,4,10,12];
y=1:15;
[fitobject,gof]=fit(y',x','gauss2');
figure
bar(y,x)
hold on
plot(fitobject,y,x)

but 1) it's unclear whether this is the best way to do this, and 2) I don't know how to extract the variance from this gaussian.

Thoughts? Thanks in advance!

Answer 1

[Star Strider]

One approach —

x=[8,11,8,10,8,21,26,33,28,8,17,7,4,10,12];
y=1:15;
[fitobject,gof]=fit(y',x','gauss1');
figure
bar(y,x)
hold on
plot(fitobject,y,x)

fitobject

fitobject = 
     General model Gauss1:
     fitobject(x) =  a1*exp(-((x-b1)/c1)^2)
     Coefficients (with 95% confidence bounds):
       a1 =       24.61  (17.1, 32.13)
       b1 =       7.828  (6.673, 8.983)
       c1 =       4.645  (2.954, 6.335)

Mean = fitobject.b1

Mean = 7.8278

Sigma = fitobject.c1

Sigma = 4.6447

Variance = fitobject.c1^2

Variance = 21.5732

Using ‘Gauss2’ is likely inappropriate here.

Experiment to get different results.

.