2 of 20

About Me

All my university work was in mathematical physics (grad, undergrad)
Then became a high school physics teacher
Then happened into spatial data science

Live in the Washington DC area
I have two little kids - a three year old and a one month old

3 of 20

Past Work - Data scientist at CARTO

Data scientist at “geospatial in the cloud” platform CARTO
Built the beta version of CARTOframes, a Python package for spatial analysis for data scientists
Helped build CARTO’s first data catalog, the Data Observatory
Explored a lot of fascinating datasets, including client data, OpenStreetMap, SafeGraph, US Census, and many others

4 of 20

Currently - Data Scientist at Ipsos

Lead geospatial scientist in the Risk Analytics Division at Ipsos, a research company
Respond to humanitarian disasters using the same tools we will be using in class
Big data processing of mobility data (aka, smartphone GPS app data)
Always looking for the next best mapping tool

5 of 20

About Felix

www.felixqiang.com

CIT master@Penn & Math@Bryn Mawr
Wrote a lot of code and dealt with many types of datasets
Also had built an AR app with location data
Love to work on projects that with a visual component to it
Ask me anything about the tech!
Currently in the Bay Area
I also take photos and make films

6 of 20

Intros around the class

Introduce yourself
What sorts of problems are you interested in exploring?
Where are you located?
Favorite vacation destination

7 of 20

Where we’re going, what we’ll build

Custom applications to answer your urban planning questions
Final project goal: everyone builds an application that takes in user-defined inputs and gives them a custom answer
Make sure to talk about applications/insights/etc, not technology. Applications are the important part, not the technology that underlies it.

8 of 20

High Level Skills and Technology

9 of 20

Spatial Databases

SELECT superhero.name

FROM city, superhero

WHERE

ST_Contains(

city.geom,

superhero.geom)

AND

city.name = 'Gotham';

Databases are a super power once you learn a bit about them. We’ll learn lots of SQL, a very transferable skill

How we’ll use them

Querying data efficiently
Managing datasets with explicit relationships between them
Query geography (e.g., give me all bike share stations within 500 meters of a cafe), find superheros, etc.

10 of 20

Large, messy datasets

We will work with large and/or messy datasets such as OpenStreetMap, SafeGraph, US Census, and more, in a variety of formats.

We’ll cover

Techniques for taming these datasets
Build tools around datasets
Learn tools for storing, accessing, and analyzing datasets (e.g., AWS S3 and Athena, Google BigQuery)

11 of 20

Enough Python to be dangerous

Python is a powerful programming language with a huge community, which means that there are a lot of amazing packages.

We need Python to

Provide glue between a user interaction on a webpage and a database transaction
Give us tools for building the project you dream up (e.g., data science ecosystem)

12 of 20

API basics and HTTP requests

An API gives an interface for computers to communicate with one another. We will all build one that will communicate between the user interactions on a web page with data stored in a database.

13 of 20

Cloud Services

Our final work will all be in the cloud!

APIs will be run from computers on AWS
Data will be stored in databases on AWS
We’ll use Google Cloud for accessing data in BigQuery

14 of 20

Class Logistics

15 of 20

Lectures

Live on Zoom: Tuesdays 4:30pm - 6pm ET

Recorded: available after lecture. Links will be posted to our class Canvas page

Slide deck and lecture notes - GitHub

16 of 20

Labs

US time zone: Thursdays 4:30pm - 5:15pm ET

Asia time zone: TBD, but preferably some time on Thursday (see first week survey to give your input)

Labs will also be detailed on

17 of 20

Participation

Ask and answer questions on Piazza
Finding mistakes on GitHub or adding something new/cool you found that could help others
Presenting during Labs

18 of 20

Office Hours

Andy

TBD - waiting to see how the second Lab is scheduled
By appointment

Felix

1 of 20