lambdata
A collection of Data Science helper functions. Currently xgb_class and xgb_reg do not output as intended and are being investigated for potential issues.
Download
https://test.pypi.org/project/lambdata-mpharm88/
What is it?
lambdatas_mpharm88 is a Python package that lets you reach your model baseline as quickly as possible. This means that when you use lambdata_mpharm88 your delivering initial observations faster to interested parties. This package also helps you find out how hard it will be to beat your mean baseline and provides functionality for preprocessing your data frame too. Lambdata_mpharm88 seeks to deliver quick, meaningful results when time constraints don't allow for a full investigation into the data.
Main Features
Here is a list of what lambdata_mpharm88 does well:
- Changing a date column into ５ new columns including:
- Pandas date time column
- Year column
- Month column
- Day column
- Season column
- Performing a train/test split using your original data frame, target (y), and features. This function will give you 8
separate dataframes including:
- X training set
- y training set
- X validation set
- y validation set
- X testing set
- y testing set
- X cross validation set
- y cross validation set
- Shapes of all dataframes outputted
- A simple XGBClassifier including an Ordinal Encoder and Simple Imputer that out puts:
- Training ROC/AUC Score
- Testing ROC/AUC score
- Mean baseline of majority class in target column
- A simple XGBRegressor including an Ordinal Encoder and Simple Imputer
- Training RMSE
- Testing RMSE
- Mean baseline of target column
Dependencies
License
Documentation
Currently, this read me is the only documentation available. If you would like to join the party and help build this package, feel free to create a pull request with your changes. All feedback is appreciated, and together we can turn lambdata_mpharm88 into a tool that's used by people all over the world.