PyData London - 62nd meetup

Venue: 1 Angel Lane, EC4R 3AB

NOTE: A valid photo ID is required by building security. You MUST use your full real names on your meetup profile, otherwise, you will NOT make it on the guest list!

Tickets are assigned through a lottery draw about 1 week before the event.

If your RSVP status says "You're going" you will be able to get in. No further confirmation required. You will NOT need to show your RSVP confirmation when signing in.

If you can no longer make it, please unRSVP as soon as you know so we can assign your place to someone on the waiting list.

---------------------------------------------------------------------------------------------

CODE OF CONDUCT: https://numfocus.org/code-of-conduct

This event follows the NumFOCUS code of conduct, please familiarise yourself with it before the event. Please get in touch with the organisers with any questions or concerns regarding the code of conduct.

---------------------------------------------------------------------------------------------

As always, there'll be free food & drinks, generously provided by our host, Man Group.

Main Talks

Cheuk Ho - Are you supporting the right politician? - Graph Visualization of Voting Data

It’s time for the election, a major of people will vote for the politician according to the political parties they affiliated. Do you really know the standpoint of politician that you vote for on different issues? Do you trust the politician that you vote for is representing you in the council/parliament? Have you been monitoring their votes in the council/parliament on different issues? In this talk, we will be identifying real ‘parties’ by voting records rather than by stated affiliation using a knowledge graph. Everyone can do it as well (with their own data), we will walk you through how to do it and hopefully, you will find the representatives that you vote for are doing the right things for you (or maybe not).

Oli Cairns - Weight of Evidence Binning

Weight of evidence binning is a feature engineering strategy that can allow logistic regression classifiers to better accommodate non-linear feature dependencies, without extensive manual feature generation, or a loss of model interpretability. It is commonly used in the lending industry to quickly develop sparse, interpretable credit-risk models, and can be useful in other domains as a benchmark, or in cases where model interpretability is particularly important. In this talk, I will give an overview of the methodology, and a brief theoretical justification. I will also demonstrate how it can be implemented in Python, and used to close some of the performance gap between unbinned logistic regression, and popular decision tree ensemble libraries such as LightGBM. I will also discuss some limitations of the modelling strategy, and approaches that can be used to overcome some of these.

Lightning Talks

Michael Grazebrook - TatSu - using grammar with Python

Tatsu lets you write an EBNF grammar and use it in Python. The Python equivalent of lex & yacc, bison or similar C/C++ tools. Useful for complexly structured information such as a programming language or computer output not designed as a data format.

Puneet Thukral - Visualization of Data using Latex and Python

A brief intro into how to generate print-ready reports with Python.

Logistics

Doors open at 6.30 pm (get there early as you have to sign-in via building security), talks start at 7 pm, drinks from 9 pm in the bar. We normally have >200 folks in the room so there's plenty of people to discuss data science questions with!

Please unRSVP in good time if you realise you can't make it. We're limited by building security on the number of attendees, so please free up your place for your fellow community members!

Follow @pydatalondon (https://twitter.com/pydatalondon) for updates and early announcements.