I’m taking a EDx course in conjunction with the Portland Python Pirates group. The best of both worlds – an on-line learning with a study group of other local people. Stay tuned!
Next Step: The MITx Python Course
Category: Data Analysis
Air Quality Index (AQI): Portland, Oregon
The air quality around Portland with numerous nearby fires was between Very Unhealthy and Hazardous using the US EPA AQI scale.
Which was quickly confirmed by stepping outside.
As Time Goes By: Coronavirus
Here is my third presentation in the coronavirus series, where I focus on and summarize six aspects of the situation:
- Coronavirus case load in Italy, US, New York state and Oregon
- Perspective of Public Health
- Viewpoint of the general public
- Collapse of Business
- Review of President Trump’s strengths
- Leadership in the US
Leading to my conclusion:
- The next 6-12 months is unpredictable. The virus remains constantly contagious. While people’s behavior stopped the initial peak. Currently the variables of peoples’ behavior of mask wearing and social distancing, with conflicting messages from American leadership is simply unpredictable.
My “Coronavirus Overview”
We see many great articles about specific topics about the coronavirus situation. I was having trouble fitting all the details together to form an overview of the situation I could understand. So I researched, and then put together this video.
- Combining the initial “keep the peak down” chart
- John Hopkins University’s great work on tracking case information
- US/Other countries & Oregon/Other states
- Timelines over the next year
- Some components of the situation
Hopefully, you will find the video helpful. I had a lot of fun making it!
Coronavirus: Update to 3/26 Projections
In late March there was a lot of talk about will we open the country on Easter? How many people will have the virus then?
I researched and estimated how many people in Oregon would test positive for coronavirus and how many would die by Easter. Healthcare stats are a bit different than modeling future sales and capacity planning! (See my blog on 3/26 for the initial information.) As Easter was yesterday, I revisited my projections against actual.
NET:
- Number of cases: I over-estimated (1,371 actual vs 2,862) As the growth rate dropped from around 15% per day to 4-8% for the last week. I had a constant 15% growth rate. Everyone stayed home!
- Number of deaths: I was very close (52 actual vs 51) As my predicted growth rate matched actual until the last week. To be blunt, it takes awhile for one to die. So my model expected most people diagnosed in the last week to still be alive.
Unemployment Rate? With 6.648M More: Answer: ~9.6%
As was reported this morning by CNN – another 6.648M Americans are newly unemployed. My mind quickly went to, “What is the updated unemployment percentage?” As the newscast did not mention this number.
- The federal government doesn’t calculate it weekly, just monthly
- So I approximated it at 9.6%
Assumptions in the approximation:
- No one unemployed the prior week found a job
- Without seasonally adjusted numbers calculations
- Sources:
- https://www.dol.gov/ui/data.pdf (yellow highlighted number)
- https://www.bls.gov/news.release/empsit.a.htm (blue highlighted numbers)
Question: With 3.283M Newly Unemployed, What is the Unemployment Rate? Answer: 5.5% Up From 3.5%
In the news stories yesterday – we were all shocked at the 3.283M Americans newly unemployed. My mind quickly went to, “What is the updated unemployment percentage?” As the newscasts did not mention this number.
The federal government doesn’t calculate it weekly, just monthly.
So I approximated it at 5.5%.
Assumptions in the approximation:
- No one unemployed the prior week found a job
- Without seasonally adjusted numbers calculations
Sources:
https://www.dol.gov/ui/data.pdf (yellow highlighted number)
https://www.bls.gov/news.release/empsit.a.htm (blue highlighted numbers)
Video: Coronavirus Review & Update
There is a lot of information flying around about Coronavirus, so I decided to dig into the data stories. Net:
- US will soon be the epicenter – country with most number of cases
- New York has the same number of cases/1 million citizens as Italy
- 1250-1600 cases
- Louisiana and Washington state about the same: ~350 cases/1 million
- US average is about 200 cases/1 million
- Oregon has far fewer cases at ~65 cases/1 million
- Oregon should see cases increase by 10x by Easter
- with relatively low daily growth rate of 22%
- Oregon should see cases increase by 10x by Easter
YouTube video: Coronavirus Review & Update
Data Story: Airbnb Booking Comparison
Our Airbnb bookings in early February seemed a little less than usual for mid-winter. So I compared our bookings for February and March to the Airbnb’s around us.
Net: We’re right in the middle of pack.
Click on the graphic below to my telling of this story on YouTube.
Wine Spectator, By the Numbers: 2019
Sometimes I geek out on charts… like this over-done one by Wine Spectator.
Basic Conclusions
- They tasted a lot of wine in 2019: 15,111 bottles (“All Wines” top left)
- There is a tremendous amount of great wine available
- The Best Wines – wine rated 90 or above
- Four-way Tie: California, France, Germany and Oregon
- California best: Cabernet & Syrah varietals
- France best : Champagne & Burgundy regions
- Germany and Oregon have no detail
- Four-way Tie: California, France, Germany and Oregon
- The Worst Wine –
- Anything from US states not in CA, NY, OR or WA
What does it say to me?
- Done Well:
- A global summary of wine quality on one page
- Color coding good: Green-best, red & yellow – worst
- Could use improvement:
- They didn’t provide their own summary!
- Data isn’t quite comparable as originally appears
- California’s detail is the grape varietals
- Other states have no detail
- Nations are by regions rather than varietal
- Liked Best:
- They kept careful track of their tasting results
- Oregon wines are as good as historical best, and
- Are price competitive
- With the note that I live in Portland, Oregon!
Josh's Blog 