km-git-acc / dbn_upper_bound

Computational effort to upper bound the de Bruijn-Newman constant as part of a Polymath project
Other
13 stars 12 forks source link

Ht eval using mpmath #27

Closed km-git-acc closed 6 years ago

km-git-acc commented 6 years ago

This branch has 3 changes.

Main readme file has been updated with latest developments, and now also contains a section to get someone new quickly started.

mputility.py has removed dependencies on the constants.py file. It uses the inbuilt pi function from mpmath which is available for any desired precision.

There is also a sample file sample_afe_calc.py showing how large scale computations with Ht_AFE (approx functional eqn) can be done.

Please review and merge.

sujitnair commented 6 years ago

Oops sorry. Just noticed this @km-git-acc

Let me go over it. Next time, you can directly assign a PR to me. That way I will get an email notification.

sujitnair commented 6 years ago

@km-git-acc

I pushed some changes to sample_afe_calc.py. Please have a look at it.

pylint was complaining a lot :( I also added os.path.join to make sure the output is going to the right folder.

Can you run pylint --rcfile=pylintrc dbn_upper_bound/python/mputility.py

The score is pretty low -1.57. Can you make the recommended changes and get the pylint score up? We should target to have at least 8 on this.

Also, how are we doing a sanity check on the computation?

sujitnair commented 6 years ago

@km-git-acc I also noticed that sometimes your indentation is using five spaces rather than four. You might want to look at your IDE/editor settings.

km-git-acc commented 6 years ago

Well, I use notepad while working, so there could be different indentations but it should be more or less the same. On the PI thing, I just did a find replace on realizing mpmath has a pi function. Not sure how utility.py would react on seeing the mpmath floating numbers, so this seemed easier.

Sanity check was done manually in excel using the actual roots of zeta from Odlyzko's webpage. I am quite comfortable in excel and this seemed easier. But it may be possible to write a script for this as well.

km-git-acc commented 6 years ago

Also, all the changes seem fine in sample_afe_calc, so merging this branch. I haven't worked with pylint before but will try it out.

sujitnair commented 6 years ago

@km-git-acc

Oops. mputility.py wasn't ready to be merged yet :(

sujitnair commented 6 years ago

@km-git-acc

You can check out the free version of PyCharm https://www.jetbrains.com/pycharm/download/

It makes developing in Python a breeze.