Blogs

This is a little side-project I have worked on during one weekend. Full story can be found here Project itself is here Project code can be found here

Christmas came early this year! Woohoo! Happy to officially announce NBAr 2.0 update (that is available on github for some time now…) List of significant changes: NBA’s api became really unstable if it’s called without referall, so it the readr::read_lines had to be replaced with httr::GET Old code sometimes looked really bad and I felt bad because of it To keep it consistent, most of the dependencies is on Tidyverse packages functions_and_argument_names_are_finally_in_snake_case ! ... Read more …

Introduction This blog post is definately not about love, I can assure you that. I have to admit to myself that after being fanatic R user for more than 3 years, it’s time to learn something new and add cool stuff to my little green toolbox. Anytime I needed to do stuff with data outside of databases I was launching rstudio and was ready to go in no time with well known and comfortable tidy libraries, pipelines and Rprojects. ... Read more …

Couple of days ago I had to find a way to visualize in R cumulative number of unique customers over period of time. Having data with dates and customer ids ready I felt pretty confident that I will finish the task in seconds and most of the time will be eaten by loading tidyverse package. Damn I could have felt the coffee break LOOMING behind my screen. Little did I know back then but apparently there is no function yet to do just that one task. ... Read more …

About series Introduction Splines Linear spline Quadratic spline Cubic spline Example Local Regression (LOESS) Generalized Additive Models (GAMs) Training GAMs About series So I came up with an idea to write series of articles describing how to apply machine learning algorithms on some examples of NBA data. Got to say that plan is quite ambitious - I want to start with machine learning basics, go through some common data problems, required preprocessing tasks and obviously, algorithms. ... Read more …

About series Linear Regression Methods of achieving the best fit Ordinary Least Squares - OLS One predictor More predictors (matrix-based formula) Stochastic Gradient Descent method Assumptions Linear relationship Correlation Homoscedasticity of error terms Normal distribution of error terms Autocorrelation of error terms No Multicollinearity Outliers High leverage observations Scaling variables Building a model - example Linear Regression with mlr package Correlation based feature selection First basic model Model with excluded/scaled observations Conclusion About series So I came up with an idea to write series of articles describing how to apply machine learning algorithms on some examples of NBA data. ... Read more …