Meesum Qazalbash

@fehiepsi Power Law is not in SciPy. Can you guide me for the Unit Tests?

@fehiepsi unfortunately I couldn't find much on internet too. I have seen many papers in astrophysics using these distributions. Therefore, I did the derivations. If you like to see, I...

`test_sample_gradient` and `test_log_prob_gradient` are failing only in case of `alpha=-1.0`. The gradient is coming out to be `jnp.nan`. I can not understand why, so I have pushed the changes for...

Your suggestion worked and the gradient is some float now! But values are still diverging. I have rechecked my derivations and implementation a couple of times. You can double-check it!

@InfinityMod, I will look into it.

@fehiepsi test failure related to gradients of `DoublyTruncatedPowerLaw` and `LowerTruncatedPowerLaw` are fixed. Please feel free to review it.

@fehiepsi They are still failing!

@InfinityMod The only problem left is the failing of test cases, can you look into it?

Can you change the base of your PR to `Qazalbash:powerlaw-dist`?

We should wait till #1860 gets merged to `master` then we will merge the `master` with my branch.