I would like to dedicate this post to Prof. Gleen Crellin, who 4-years-ago put trust in me, who does not have any technical background except a 10 credit summer database certificate, and gave me my very first database development job. Prof. Crellin lost his battle to cancer in May 2017. My thoughts go to his family.
Prof. Glenn Crellin
It all begins with someone who puts trust in you. After the 11-week Summer certificate program and right before I started graduate school program, I was itching to find a real life project and use all that knowledge I leaned. I love real estate (and you already know that), so I naturally looked up real estate research centers at my home school University of Washington. That’s how I found Runstad Center for Real Estate Studies. There was no job posted – whatever the worst they could say is no – but I contacted the program manager anyway. She forwarded my inquiry to Glenn, and that’s how my life started to change.
First email I sent to Runstad center’s program manager.
I have to say, I did not do a great job on this first project. I learned a ton about the importance of requirements gathering and how to collaborate with existing technical team. It was hard – I didn’t really know what I was doing. I realized the project was in a downward spiral and I reached out to Glenn. He put his hand together, looked at me, and said: “I know. I wanted to give you this chance and I know you tried hard. It’s alright.”
I want to remember this for the rest of my career. Thank you Glenn.
Hello world! It’s been a long time since I updated this blog. Now seems like a good time.
I finally registered for Tableau Public, check out my fascinations here. Obviously this is going to be mostly neighborhood / real estate related.
It was kinda fun to go back in to data.seattle.gov to look for treasure again. I started with a list of page visit data for seattle.gov. I was positively impressed to see that Socrata, a local business, now supplies a direct OData streaming on the site!
Bounce Rate to Exit Rate in Seattle.Gov/Neighborhoods pages
I was not surprised to see that there is a slight negative correlation between median unique page view to median bounce rate. If I go on the site and open one of the links, read it for the first time, my view would be counted as a unique view. I would insinuate that the more unique views a page gets, the higher chance that a viewer is actually reading the page (and not bouncing, which is leaving the page within 2-3 seconds).
From this scatter plot, I learned that people probably dont get their school information here – the “major-institutions-and-school” page groups has very low unique views but a ~45% bounce rate. It also appears that the ‘equitable-outreach-and-engagement” gets most the attention here, pretty high unique views and very low bounce rate. Curious? Let’s give them some more page views here.
I haven’t been able to talk much about real estate lately.. perhaps it’s because I have been sooo busy with MSIM (click to see program description in case you are wondering what it is). I want to show off some of the course assignments I have done lately.
My teammate and I made this visualization with D3:
We used GeoJson for a US map with state boundaries, and the blue-purple shows the median household income in that state. The lighter the purple, the lower the median income. The yellow circle shows the number of violent crime per 100,000 people in that state. The bigger the circle, the more crime incident per 100,000 people. You can change year on the slider bar. Data range is 2007 to 2013. If you hover your mouse over a state, detail data for crime and income will show up on the bottom right.
Below is the interactive map I made 🙂
Visualization is here: http://students.washington.edu/feig/
To download all the files (html, css, json, csv) to try it yourself, go to http://bit.ly/17eeV1U
I also shared it on github: https://github.com/rosefei/d3.git
A good and slow tutorial for a starter. Actual material starts at 15min.