LibNotMR gives different results from libmr.MR

[nikita-petrashen]

[nikita-petrashen]

code: `import libmr from matplotlib import pyplot as plt

np.random.seed(2)

taillen = 20 not_mr = LibNotMR(taillen) mr = libmr.MR()

tail = np.random.randn(100) test = np.arange(-5, 10, 0.01)

not_mr.fit_high(torch.tensor(tail))

mr.fit_high(tail, taillen)

plt.plot(test, mr.w_score_vector(test), label='libmr') plt.plot(test, not_mr.w_score(test), label='scipy') plt.legend() plt.show()`

[nikita-petrashen]

I don't know what is the origin of this, but actually LibNotMR plots look more reasonable