Welcome to the 30th PyData Cambridge virtual meetup!

This is a virtual meetup. We will use zoom, login details published closer to the date.

Thanks a lot to our sponsors: NumFOCUS, ARM, Fetch.ai and the Raspberry Pi foundation.

Agenda

19:00 - Introduction
19:15 - Using perceptual data to move towards the future of virtual reality - Gyorgy Denes
19:45 - Interval
19:50 - Analyzing Complex Survey Data Using Python - Mamadou S. Diallo
20:20 - End
Note: this meetup will be recorded.

Code of Conduct

PyData is dedicated to providing a harassment-free event experience for everyone, regardless of gender, sexual orientation, gender identity, and expression, disability, physical appearance, body size, race, or religion. We do not tolerate harassment of participants in any form.

The PyData Code of Conduct governs this meetup. ( http://pydata.org/code-of-conduct.html ) To discuss any issues or concerns relating to the code of conduct or the behavior of anyone at a PyData meetup, please contact NumFOCUS Executive Director Leah Silen (leah@numfocus.org) or organizers.

Talk 1

Title:
Using perceptual data to move towards the future of virtual reality

Abstract:
Rendering photo-realistic augmented reality (AR) and virtual reality (VR) content has once again become the Holy Grail of computer graphics and video games, with many anticipating wide-spread use within the next decade. However, producing perceptually perfect frames for AR and VR headsets is prohibitively expensive for current and upcoming generations of GPUs. Current video games often trade off quality for performance, resulting in a suboptimal experience with pixelated content and the use of bulky headsets. Novel approaches attempt to trick human observers using techniques such as reprojection, temporal multiplexing, adaptive and foveated rendering. The major obstacle before widely deploying these techniques in production is to characterise the quality-performance trade-off, which ultimately relies on understanding the preferences of the final consumer: us humans. In the talk I will discuss the general process of collecting and modelling human observer data in psychophysical experiments, focusing on the challenges of efficiently sampling the multi-dimensional space of VR content in order to estimate linear quality scores.

Talk 2

Title:
Analyzing Complex Survey Data Using Python

Abstract:
Survey samples are often selected, from finite populations, using predefined probabilistic methods. To facilitate fieldwork and keep costs under control, complex sampling designs (e.g., stratification, clustering, stage sampling, etc.) are used resulting in samples with unequal probabilities of selection. Selecting, estimating, and analyzing data from such complex designs require advanced sampling techniques. I developed a Python package named samplics which implements sample size calculation, sample selection, population parameter estimation and small area prediction. This package will allow Python users to easily work with survey data without being required to use R or other non-Python survey sampling packages. samplics is designed for the analysis of large and complex surveys and the production of official statistics. The Python package samplics is in beta version and not yet ready for production, but it is in active development and can already be used for many survey analyses. In this talk, I will show the functionalities of samplics.