Weird fitting result from using 'ksdensity'

I have a histogram of frequency distribution (pic 1), theoretical there could be more than one peak, so I used ksdensity for fitting. But I got a reallly weird fitting.(pic 2)

code I used:

figure
x = [Tracks.norm_amp];
histogram(x,'Normalization','probability')
[f,xi] = ksdensity(x,'kernel','normal','support','positive');
hold on
plot(xi,f)

It wil be great if someone can help me! Thanks a lot!

Answer 1

[Jeff Miller]

Try

histogram(x,'Normalization','pdf')

[Yiran Chang]

Worked! I really appreciate your help!