search

Shaneul / MLE

Code for fitting to degree distributions of complex networks using maximum likelihood estimators
3 stars 1 forks source link

MLE issue on exponential fitting #1

Open rmbw51 opened 2 months ago

rmbw51 commented 2 months ago

MLE works very well but I found a problem with the fitting of an exponential distribution. I have an example where where MLE correctly identifies and plots well, but the exponent calculated is way too big, fit should be something around 0.4 (fitted in 2-3 ways) while the MLE gives 2.4... any idea of why? Many thanks PS if you need I'll attach the network.

Shaneul commented 2 months ago

Hi,

The code uses (1/\lambda) as the exponential distribution exponent, so perhaps you just need to invert your answer?

rmbw51 commented 2 months ago

ok, thanks