We typically have a main presentation or a series of lightning talks, followed by discussion and Q&A. There is a diversity of domains and experience levels represented, so come with your questions and be prepared to talk about how you use Python!

This will be an online meeting - please join the meetup at the link below. Please note that this link may update and updates may appear in the discussion section below - so scroll down if you have technical difficulties.

The presentations will start after 7, yet feel free to join starting 6:30.

We will have 2 presentations:
Talk 1: Rencheng Dong will discuss Making Interactive Maps with Python using “Folium”
Talk 2: Erik Case will discuss Hyperparameter tuning with hyperopt

Below are details about the talks:

Talk 1: Rencheng Dong will discuss Making Interactive Maps with Python using “Folium”

Folium enables users to manipulate the data in the Python ecosystem, then visualize it in a Leaflet map. This workflow will be demonstrated through an example of mapping COVID-19 cases in Austin. See: https://renchengdong.github.io/austinCOVIDMap/.
The steps involve: 1) Obtain data through automated web scraping. 2) Integrate data using GeoPandas. 3) Map geospatial data with Folium. 4) Implement handy features like highlight, tooltip and search functions.

Speaker Bio: Rencheng Dong is a PhD student in the Center for Subsurface Modeling at UT Austin. He is working on numerical modeling of subsurface reactive flow processes. In his free time, Rencheng enjoys swimming, camping and working on small projects.

Talk 2: Erik Case will discuss Hyperparameter tuning with hyperopt

Hyperparameter tuning often involves methods like grid search, which are computationally expensive and memory-less. The Hyperopt package in Python enables an iterative Bayesian search through the hyperparameter subspace, arriving at faster solutions for optimal hyperparameter settings. I will do a high level overview of Hyperopt, and if time permits, demo a grid vs Bayesian approach.

See the library on Github: https://github.com/hyperopt/hyperopt

Speaker Bio: I come from the social sciences (Econ and Poli-Sci) and have been working for 10 years as an applied statistician/senior data scientist in finance and marketing. I currently work for F5 networks out of Seattle and manage a team of data scientists, where we have used Hyperopt in practice for our modeling efforts. My pet passion/white whale is time series - particularly dealing with time series + cross sectional data (panel data).