$ cat luca.json
{
"work": [
"π» Machine Learning Engineer at www.xtreamers.io",
"π Computer Science and Statistics TA at www.unimi.it"
],
"open-source": [
"functime-org/functime"
],
"interests": [
"π Time series forecasting",
"π· Machine learning",
"π¨ Data visualisation"
]
}
- π¬
pbg-functime
: Scalable time-series forecasting with @functime-org @PythonBiellaGroup - π¬
pbg-polars
: @pola-rs marathon @PythonBiellaGroup - π¬
pydata-global-2023-functime
: @pola-rs and @functime-org talk at @pydata - π¬
kaggledays-functime
: Scalable time-series forecasting with @pola-rs - π¬
pyconpt-polars
: I knowPolars
is fast, but my data pipelines are written inpandas
! - π¬
pyconit-polars
:Polars
: is the great dataframe showdown finally over? π»ββοΈ v πΌ - π¬
pyconit-rust-is-easy
: Rust is easy, or what Pythonistas can learn to improve on their code π - π¬
pymi-pyo3
: how to usematurin
andPyO3
to build python packages with a Rust core π¦ - π¬
pymi-polars
: get stared withPolars
, a fast dataframe library with an embarrassingly parallel execution engine, written in Rust π»ββοΈ - π¬
pymi-cookiecutter
: get started withcookiecutter
, a library to build coding project templates πͺ - πΆ
awesome-pandas-alternatives
: explore dataframe libraries alternative topandas
.