2022
- Physionet 2019 preprocessing
- Physionet 2012 preprocessing
- UCR time series classification data
- Tabular classifier with embeddings
- Polars cookbook
- Neural adaptive regression splines
Some methods use Spark for data processing. To install Spark on MacOS, run brew install apache-spark
. You will need to install the python modules pyspark and findspark. All required python modules are in requirements.txt.
You can load (and reload) the modules like utils.py
using the snippet below:
# Add the parent notebook folder to the modules search path
parent = '../'
if parent not in sys.path: sys.path.append(parent)
import importlib
import utils
importlib.reload(utils)
utils.py includes generic utility functions to read/write parquet datasets, impute missing values, summarize dataframes, pad sequences, add styles to plots, initialize spark, and transform design matrices.