Use Positron Assistant with GitHub Copilot
The newest monthly release of Positron has a revamp of the integration between GitHub Copilot and Positron Assistant.
Julia Silge
Recent content on Julia Silge
Latest posts
Explore #TidyTuesday literary prizes with Positron’s Data Explorer
The newest monthly release of Positron delivers some fresh new features for the Data Explorer.
Release an R package with Positron
See how you can use Positron, a new, next-generation data science IDE, for R package development tasks and releasing a new version of an R package.
Positron in action with #TidyTuesday orca encounters
Get to know Positron, a new, next-generation data science IDE, using this week’s Tidy Tuesday data on encounters with orcas.
Educational attainment in #TidyTuesday UK towns
Let’s walk through the ML lifecycle from EDA to model development to deployment, using tidymodels, vetiver, and Posit Team.
Changes in #TidyTuesday US polling places
Let’s use summarization and visualization to explore how the numbers of polling places have changed in the United States.
Empirical Bayes for #TidyTuesday Doctor Who episodes
Which writers of Doctor Who episodes are rated the most highly? Let’s use empirical Bayes to find out.
Logistic regression modeling for #TidyTuesday US House Elections
Today is Election Day in the United States, so let’s use logistic regression modeling to explore vote share in US House elections.
Topic modeling for #TidyTuesday Taylor Swift lyrics
Learn how to fit and interpret an unsupervised text model for all of Taylor Swift’s ERAS.
Where are #TidyTuesday haunted cemeteries compared to haunted schools?
Use tidy log odds to compare which US states are more likely to have haunted cemeteries or haunted schools.
How often does Roy Kent say "F*CK"?
He’s here, he’s there, he’s every f*cking where, and we’re finding bootstrap confidence intervals.
Evaluate multiple modeling approaches for #TidyTuesday spam email
Use workflowsets to evaluate multiple possible models to predict whether email is spam.
Classification metrics for #TidyTuesday GPT detectors
Learn about different kinds of metrics for evaluating classification models, and how to compute, compare, and visualize them.
What tokens are used more vs. less in #TidyTuesday place names?
Let’s use byte pair encoding tokenization along with Poisson regression to understand which tokens are more more often (or less often) in US place names.
Predict the magnitude of #TidyTuesday tornadoes with effect encoding and xgboost
How well can we predict the magnitude of tornadoes in the US? Let’s use xgboost along with effect encoding to fit our model.
Tune an xgboost model with early stopping and #TidyTuesday childcare costs
Can we predict childcare costs in the US using an xgboost model? In this blog post, learn how to use early stopping for hyperparameter tuning.
Deploy a model on AWS SageMaker with vetiver
Learn how to train and deploy a model with R and vetiver on AWS SageMaker infrastructure.
Use OpenAI text embeddings with #TidyTuesday horror movie descriptions
High quality text embeddings are becoming more available from companies like OpenAI. Learn how to obtain them and then use them for text analysis.
Resampling to understand gender in #TidyTuesday art history data
Artists who are women are underrepresented in art history textbooks, and we can use resampling to robustly understand more about this imbalance.
To downsample imbalanced data or not, with #TidyTuesday bird feeders
Will squirrels will come eat from your bird feeder? Let’s fit a model both with and without downsampling to find out.
High cardinality predictors for #TidyTuesday museums in the UK
Learn how to handle predictors with high cardinality using tidymodels for accreditation data on UK museums.
Delete all your tweets using rtweet
Worried about how a certain social media platform is going and want to start removing yourself? Learn how to delete all your tweets.
Find high FREX and high lift words for #TidyTuesday Stranger Things dialogue
New functionality in tidytext supports identifying high FREX and high lift words from topic modeling results.
Predict the status of #TidyTuesday Bigfoot sightings
Learn how to use vetiver to set up different types of prediction endpoints for your deployed model.
Use Docker to deploy a model for #TidyTuesday LEGO sets
After you train a model, you can use vetiver to prepare a Dockerfile and deploy your model in a flexible way.
Sliding windows for #TidyTuesday rents in San Francisco
The slider package provides support for flexible sliding window aggregation, and we can use these kinds of sliding windows to analyze rents over time.
Three ways to look at #TidyTuesday UK pay gap data
Use summarization, a single linear model, and bootstrapping to understand what economic activities involve a larger pay gap for women.
Use resampling to understand #TidyTuesday drought in TX
The spatialsample package is gaining many new methods this summer, and we can use spatially aware resampling to understand how drought is related to other quantities across Texas.
Predict #TidyTuesday NYT bestsellers
Will a book be on the NYT bestseller list a long time, or a short time? We walk through how to use wordpiece tokenization for the author names, and how to deploy your model as a REST API.
Handling model coefficients for #TidyTuesday collegiate sports
Understand how much money colleges spend on sports using linear modeling and bootstrap intervals.
Poisson regression for #TidyTuesday counts of R package vignettes
The tidymodels framework provides extension packages for specialized tasks such as Poisson regression. Learn how to fit a zero-inflated model for understanding how R package releases are related to number of vignettes.
Inference for #TidyTuesday aircraft and rank of Tuskegee airmen
The infer package is part of tidymodels and provides an expressive statistical grammar. Understand how to use infer, and celebrate Black History Month by learning more about the Tuskegee airmen.
Predict ratings for #TidyTuesday board games
Use custom feature engineering for board game categories, tune an xgboost model with racing methods, and use explainability methods for deeper understanding.
Text predictors for #TidyTuesday chocolate ratings
Get started with feature engineering for text data, transforming text to be used in machine learning algorithms.
Topic modeling for #TidyTuesday Spice Girls lyrics
Learn how to train, explore, and understand an unsupervised topic model for text data.
Predicting viewership for #TidyTuesday Doctor Who episodes
Using a tidymodels workflow can make many modeling tasks more convenient, but sometimes you want more flexibility and control of how to handle your modeling objects. Learn how to handle resampled workflow results and extract the quantities you are interested in.
Spatial resampling for #TidyTuesday and the #30DayMapChallenge
Use spatial resampling to more accurately estimate model performance for geographic data.
Predict #TidyTuesday giant pumpkin weights with workflowsets
Get started with tidymodels workflowsets to handle and evaluate multiple preprocessing and modeling approaches simultaneously, using pumpkin competitions.
Multiclass predictive modeling for #TidyTuesday NBER papers
Tune and evaluate a multiclass model with lasso regulariztion for economics working papers.
Dimensionality reduction for #TidyTuesday Billboard Top 100 songs
Songs on the Billboard Top 100 have many audio features. We can use data preprocessing recipes to implement dimensionality reduction and understand how these features are related.
Fit and predict with tidymodels for #TidyTuesday bird baths in Australia
In this screencast, focus on some tidymodels basics such as how to put together feature engineering and a model algorithm, and how to fit and predict.
Modeling human/computer interactions on Star Trek from #TidyTuesday with workflowsets
Learn how to evaluate multiple feature engineering and modeling approaches with workflowsets, predicting whether a person or the computer spoke a line on Star Trek.
Predict housing prices in Austin TX with tidymodels and xgboost
More xgboost with tidymodels! Learn about feature engineering to incorporate text information as indicator variables for boosted trees.
Supervised Machine Learning for Text Analysis in R is now complete
Our new book in the Chapman & Hall/CRC Data Science Series is now complete and available for preorder!
Tune xgboost models with early stopping to predict shelter animal status
Early stopping can keep an xgboost model from overfitting.
Use racing methods to tune xgboost models and predict home runs
Models like xgboost have many tuning hyperparameters, but racing methods can help identify parameter combinations that are not performing well.
Predict which #TidyTuesday Scooby Doo monsters are REAL with a tuned decision tree model
Which Scooby Doo monsters are REAL?! Walk through how to tune and then choose a decision tree model, as well as how to visualize and evaluate the results.
Create a custom metric with tidymodels and NYC Airbnb prices
Predict prices for Airbnb listings in NYC with a data set from a recent episode of SLICED, with a focus on two specific aspects of this model analysis: creating a custom metric to evaluate the model and combining both tabular and unstructured text data in one model.
Class imbalance and classification metrics with aircraft wildlife strikes
Handling class imbalance in modeling affects classification metrics in different ways. Learn how to use tidymodels to subsample for class imbalance, and how to estimate model performance using resampling.
Partial dependence plots with tidymodels and DALEX for #TidyTuesday Mario Kart world records
Tune a decision tree model to predict whether a Mario Kart world record used a shortcut, and explore partial dependence profiles for the world record times.