the data diary

Stay in touch

2024 POTUS Model Retrospective

What went well, what could be improved, and where to go from here

Um, Factually

A power ranking for the title of most pedantic nerd on Dropout’s Um, Actually

A forecast of the 2024 presidential election

The methodology behind the model

No, Biden shouldn’t be replaced as the Democratic nominee

If you want to see a Democrat in the white house next year, Biden is your best bet

Churn, baby, churn

A survivor’s guide to estimating customer lifetime from discrete censored churn data

A Change to the Time for Change

How to think about (and avoid thinking about) the presidential election in 500 days

3-PVI

A new metric for measuring regional partisanship

Free lunch at the vectorization table

Reparameterizing multinomial models for better computational efficiency

Gauss’ Goal Generation

Using gaussian process models for hospital goal setting

Mr. P and the Multiverse of Madness

Estimating population results from non-representative survey samples using multilevel regression and post-stratification

Trump vs. Desantis in 2024 Republican Primary Polling

Mixed scales lead to mixed results

My 2022 Magnum Opus

Ordered categorical models for estimating Net Promoter Score: a hierarchical Bayesian implementation

Chief Twit

A causal look at how Musk’s twitter ownership has affected Tesla stock

Hierarchical Hospitals

Multilevel Models for Patient Satisfaction

Forecast Launch!

A look under the hood of the 2022 Midterm Model

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…

Introducing {nplyr}

A grammar of (nested) data manipulation

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…

Estimate your uncertainty

Modeling patient satisfaction data with empirical Bayesian methods

Practical Data Visualization Tips for Excel Users

Improving a plot, one step at a time

“30 is not Statistical”

Organizational resistance to uncertainty

Impeachment Republicans and Democracy

How republicans who voted to impeach or acquit Donald Trump differ in their votes on democracy

Introducing {workboots}

Generate bootstrap prediction intervals from a tidymodel workflow!

The Data Science Hierarchy of Needs

Why you should put a roof on your house before hiring an interior decorator

Pull Yourself Up by Your Bootstraps

Using bootstrap resamples to generate confidence intervals from non-linear models

How Popular is Joe Biden?

Building a Presidential Approval Model

The Data Diary: Year in Review

In the year since I started this blog, there’s been a lot that’s happened: I learned to use R, picked up the basics of machine learning, and moved into a new job/industry. I…

Do Voters Want Democrats or Republicans in Congress?

Building a Generic Ballot Poll Aggregator

Technical Books!

Happy (belated) Thanksgiving! This year, my family drove down to Houston for the holiday & I hosted Thanksgiving for the first time. We played lots of games and ate well -…

Diamonds are Forever

Feature Engineering with the Diamonds Dataset

Election Night: Some Closing Thoughts on the VA Governor Race

A few weeks ago I had written about the VA Governor’s race before going on vacation - in that time it seems as though Terry McAuliffe’s campaign had lost a lot of steam and…

Polling Average of the VA Governor’s Race using purrr::map functions

Rolling poll averages can be misleading in the absence of errorbars or an expected distribution of outcomes. FiveThirtyEight is currently tracking polls of Virginia’s…

RStudio’s Call for Documentation

In my job, I spend a good amount of time working in the platform setup by our survey vendor, Qualtrics. There are some pre-formatted reports that we can send on a recurring…

Catching Up (again)

Wow, it’s been quite a while! Once again, I’m catching up from a long hiatus (this time, there was about a month’s gap between posts). Although I haven’t been writing here…

Tidymodels and the Titanic

This week, I thought I’d do something a bit different. I’ve been working with & getting used to tidymodels, a suite of R packages for building machine learning models with…

The Gas Price Fallacy

As America has vaccinated its population, life has seemingly begun to return to the pre-pandemic normal. Businesses have been opening to higher levels of capacity, schools…

Blexas?

Last year, Donald Trump won Texas’ 38 electoral votes handily, earning 52% of the vote compared to Biden’s 46% (R+6). Taken on its own, this is hardly surprising - Texas has…

COVID Cases Improve with Introduction of Vaccines

As vaccines have become widely available in the US, new COVID cases and deaths have dropped significantly from winter peak, with the 7-day average for both cases and deaths…

President of the Polls

Elliot Morris wrote in his newsletter last week about the increase in the public’s opinion that the government ought to be doing more. A summary of the last twenty years of…

A Lukewarm Case for DC Statehood

Once again, I’ll be keeping this very short, as I’ve continued to primarily focus on building out the demographic/economic database. I, did, however, sneak in some time to…

Catching Up

The past few weeks have been a bit lite (read::absent) in terms of posting updates. When not taking the machine learning course, I’ve been spending quite a bit of time…

Seasonality is a Weak Predictor of Border Crossings

Last week, the Washington Post published an article postulating that the recent increase in crossings at the border cannot be attributed to Biden administration policies…

Recent Works

One of the history teachers at my highschool was known for his prolific catchphrase, “give me the stuff, not the fluff,” often uttered to students attempting to submit…

aRtwork!

This week, I did something a bit different - the artwork above was made in R! The code to create the graph is actually pretty short, only taking up 28 lines. The bulk of the…

Doug Collins Saved Raphael Warnock’s Senate Bid

I’ll be honest - prior to writing this post, I had never heard of Doug Collins (R), the third major candidate in the race for Georgia’s special senate election after Raphael…

(Kind of) Projecting the 2020 Election

I’m about 4 months too late, but to practice forecasting, I made a modest projection for the 2020 national popular vote using polling data from the two weeks leading up to…

Scorecasting

A baseball forecast that correctly predicts the winner of the 15 opening day games could be a truly accurate model, or could just be getting lucky. Over the course of 6…

Some Worthwhile Links

Rather than diving deep into a topic this week, I took a bit of a break to focus on playing catchup with the stats course I’m taking. Instead, I’ve listed out below a number…

The Min-Wage Debate

On January 26, House Democrats introduced a bill to raise the federal minimum wage from $7.25 per hour to $15 per hour by 2025. In the weeks leading up to the introduction…

Your Doctor Probably Isn’t an Idiot

There’s a great standup routine about doctors (which I spent about an hour trying to find online with no luck!) with a punchline to the effect of, “getting a positive cancer…

R, ggplot2, & plotly

Editor’s note: I’ve gone back & read my previous post, & have one general thought - yeeeesh… Everything is so… “matter of fact” in tone & doesn’t really sound like me. I’ll…

Baby Steps

Today marks the beginning of what I imagine will be a long journey, with a significant number of hiccups & frustrations along the way. The end goal I have is to build out an…

Hello World!

I’m currently setting up the website & will have an official post soon(ish).