Hierarchical Hospitals

Hierarchical Hospitals If the past year of working at a large hospital system has taught me one thing, it’s that hospitals are a Russian nesting doll of structure. Within the hospital system, there are several campuses.

Finding new wedding bops with {tidyclust} and {spotifyr}

Last November, I (finally) popped the big question and proposed! Since then, my fiance and I have been diligently planning our wedding. While we have most of the big-ticket items checked off (venue, catering, photography, etc.

Introducing {nplyr}

Data manipulation and transformation is a fundamental part of any analysis. There are excellent tools in the R ecosystem for manipulating data frames (dplyr, data.table, and arrow, to name a few).

The Math Behind workboots

Generating prediction intervals with workboots hinges on a few core concepts: bootstrap resampling, estimating prediction error for each resample, and aggregating the resampled prediction errors for each observation. The bootstraps() documentation from {rsample} gives a concise definition of bootstrap resampling:

Estimate your uncertainty

I recently picked up David Robinson’s book, Introduction to Empirical Bayes. It’s available online for a price of your own choosing (operating under a “pay-what-you-want” model), so you can technically pick it up for free, but it’s well worth the suggested price of $9.

Practical Data Visualization Tips for Excel Users

I am an avid R user and will always advocate that others use R (or another programming language) for generating reproducible visualizations. In just about every organization, however, Excel plays an important role in an analyst’s toolkit.

"30 is not Statistical"

In my role as an analyst, my team and I are required to put together reports that summarize each hospital’s patient satisfaction performance in a table. These are reviewed by our system’s executive leadership team and the hospital directors in monthly operational reviews (MORs).

Introducing {workboots}

Sometimes, we want a model that generates a range of possible outcomes around each prediction and may opt for a model that can generate a prediction interval, like a linear model.

The Data Science Hierarchy of Needs

I’ve never built a house (shocking, I know), but from far too much time spent watching HGTV, I understand the basic gist of it. You lay a foundation, setup framing and walls, route mechanical and electrical, then work on final touches like painting and decorating (to be sure, I’m hand-waiving a lot of detail away here).

Pull Yourself Up by Your Bootstraps

Note (3/14/22): This article was written prior to the release of the {workboots} package. Since the release of that package, I’ve discovered some errors with the methodology described here and would recommend instead referencing the post associated with the release.