Jeff Alstott

Interesting! Any idea how to create similar functionality with current versions of networkx? Maybe we could talk to the networkx folks.

Fixed! On Tue, Aug 17, 2021 at 10:01 AM Lorenz Hilfiker ***@***.***> wrote: > Using the current version of scipy, imports from scipy.misc are now from > scipy.special. That should...

Part of the issue here is that `bins` was not an advertised feature of `pdf` (or `plot_pdf`) to begin with; it's not docstring, for example. I'm hesitant to enable/call attention/maintain...

Thanks for using powerlaw! Sorry there's an issue. If you submit a backwards-compatible pull request I'll definitely integrate it! On Fri, Aug 19, 2022 at 2:29 AM Henry Martin ***@***.***>...

I don't understand the issue. Can you be more detailed in what you're expecting to happen, and how this is different from what you're seeing?

What you're talking about was discussed here: https://groups.google.com/d/msg/powerlaw-general/A3aI-5efAuE/wpp6EGZut7gJ @psinger implemented a proportionality constant ("shifting") capability in his fork [here](https://github.com/psinger/powerlaw/commit/7dab5e2fd7fe6d7a64d08311bf19b3f853a08581), but it didn't materialize into a fully-vetted pull request to merge...

Thanks @pyguy3. As I mentioned back in 2017, if a fully-vetted pull request is made, I'll merge it in! @psinger got one started in his fork [here](https://github.com/psinger/powerlaw/commit/7dab5e2fd7fe6d7a64d08311bf19b3f853a08581).

You might be right! 1. For discrete distributions, it looks like the code is handling things fine. Can you test this directly by making a discrete pdf made from real...

- Discrete: The CDF get calculated using the `Distribution` class `cdf` function, which accounts for `xmax`: https://github.com/jeffalstott/powerlaw/blob/6732699d790edbe27c2790bf22c3ef7355d2b07e/powerlaw.py#L782-L783 What about the PDF? - Continuous: Looks like you have found a bug!...

Thanks, Charles! I am confused: I don't see how dynamics of "approaching the limit" of `xmax` relate to whether or not `powerlaw` is chopping off probability mass for values higher...