by dvorakt | Jul 15, 2020 | data science
A polished Tableau viz showing the collapse in traffic collisions during stay-at-home orders. You can download the Tableau workbook here, read the accompanying article on Medium.com here, code and data here. You can download the Tableau workbook here, accompanying...
by dvorakt | Jan 19, 2020 | data science
Showing cabinet departures for recent presidents was one of 56 “best and weirdest” charts of FiveThirtyEight in 2019. I asked my students to replicate the chart in Tableau. See if you can tell which of the above images is the FiveThirtyEight chart, and...
by dvorakt | Jun 18, 2019 | data science
In 2007, I published the Sample Annotated Paper in Econometrics in the Journal of Economic Education (JEE). That paper illustrated the structure of an econometrics paper: sections, presentation of descriptive statistics, results etc. Recently, a couple of colleagues...
by dvorakt | Jan 30, 2019 | data science
Train, validate and test partitions for out-of-time performance take planning and thought (This piece is also at TDS.) The purpose of supervised machine learning is to classify unlabeled data. We want algorithms to tell us whether a borrower will default, a customer...
by dvorakt | Mar 29, 2018 | data science, retirement plan design
A target date fund (TDF) is a sensible and increasingly common choice for retirement saving. My own employer, Union College, relies on Fidelity or TIAA TDFs as the default investment option. I was therefore curious to investigate recent reports that Fidelity is...
by dvorakt | Mar 20, 2018 | data science
Tableau’s LOD expressions are extremely powerful. They allow users to control how Tableau aggregates data. This is particularly important when we join tables with different level of aggregation. For example, combining transaction level data with store level data...
