Power spectrum of Poisson noise with λ=2

[arcTomato]

TL;DR

Poisson noise

Dear all.
I have made the power spectrum of Poisson noise(expected value $λ=2$), and it becomes like this

I think this is not good. I don't know why the power is so high when Random variable(x axis) is 0.
I tried another expected value version ,but the result didn't change.
so I would like to know the correct shape of the power spectrum of Poisson noise.
If you can help me, please
Thank you.

 

[Baluncore]

There is a huge DC offset in your data.
It needs to average close to zero to eliminate the zero frequency power.

 

[arcTomato]

Thanks @Baluncore!
Can I just replace the power with zero?like this?

 

[Baluncore]

Yes.
I would expect a flat spectrum.

 

[arcTomato]

Thank you @Baluncore!
If you don't mind, Could you teach me why does it happen? or show me some links?

 

[Baluncore]

You use auto-scale, so the huge DC offset attenuated the data.
I do not know the structure of your input data. I expect it is randomly located positive pulses?

You should increase the number of points and the number of pulses to get better detail.
You might plot several on the same graph to see any repeated pattern.
You might accumulate the power spectrum to average out the structure of the spectrum.

Google “shot noise spectrum”.

 

[arcTomato]

yes I used this data.

I appreciate for your kindness,@Baluncore !

 

[Baluncore]

If you use an FFT to compute the spectrum you should have 2^integer input data cells to get the optimum speed of computation. I would start with 1024 points, giving 512 frequencies.