documentation goals

[ClimbsRocks]

• [ ] have an examples folder
  • [ ] basic regressor
  • [ ] basic classifier
  • [ ] advanced options
  • [ ] same, but in an ipython notebook
  • [ ] use a large dataset (https://www.kaggle.com/c/porto-seguro-safe-driver-prediction/data)
• [ ] examples to include:
  • [ ] changing model names
  • [ ] feature_scaling
  • [ ] feature_selection
  • [ ] training_params
  • [ ] quantile stuff
  • [ ] nlp
  • [ ] gs_params
  • [ ] categorical_ensembling
  • [ ] feature_learning
  • [ ] comparing multiple models
  • [ ] verbose scoring
  • [ ] X_test
  • [ ] calibrate
  • [ ] lgbm_memory_optimized
  • [ ] ensembling basically, browse through our test suite, and include a bunch of those examples

[ClimbsRocks]

have a section for Get started totally from scratch! how to install python. how to update your compilers for lightgbm. how to install fastparquet to save to parquet format (with snappy?) or tables to save to .h5. how to set up a virtualenv and navigate the cli and run the script from the command line. maybe how to install what you need to connect to a sql database.

[Sandy4321]

sounds great, where to read it?