XGBoost + Optuna

Overview

AutoXGB

XGBoost + Optuna: no brainer

  • auto train xgboost directly from CSV files
  • auto tune xgboost using optuna
  • auto serve best xgboot model using fastapi

NOTE: PRs are currently not accepted. If there are issues/problems, please create an issue.

Installation

Install using pip

pip install autoxgb

Usage

Training a model using AutoXGB is a piece of cake. All you need is some tabular data.

Parameters

###############################################################################
### required parameters
###############################################################################

# path to training data
train_filename = "data_samples/binary_classification.csv"

# path to output folder to store artifacts
output = "output"

###############################################################################
### optional parameters
###############################################################################

# path to test data. if specified, the model will be evaluated on the test data
# and test_predictions.csv will be saved to the output folder
# if not specified, only OOF predictions will be saved
# test_filename = "test.csv"
test_filename = None

# task: classification or regression
# if not specified, the task will be inferred automatically
# task = "classification"
# task = "regression"
task = None

# an id column
# if not specified, the id column will be generated automatically with the name `id`
# idx = "id"
idx = None

# target columns are list of strings
# if not specified, the target column be assumed to be named `target`
# and the problem will be treated as one of: binary classification, multiclass classification,
# or single column regression
# targets = ["target"]
# targets = ["target1", "target2"]
targets = ["income"]

# features columns are list of strings
# if not specified, all columns except `id`, `targets` & `kfold` columns will be used
# features = ["col1", "col2"]
features = None

# categorical_features are list of strings
# if not specified, categorical columns will be inferred automatically
# categorical_features = ["col1", "col2"]
categorical_features = None

# use_gpu is boolean
# if not specified, GPU is not used
# use_gpu = True
# use_gpu = False
use_gpu = True

# number of folds to use for cross-validation
# default is 5
num_folds = 5

# random seed for reproducibility
# default is 42
seed = 42

# number of optuna trials to run
# default is 1000
# num_trials = 1000
num_trials = 100

# time_limit for optuna trials in seconds
# if not specified, timeout is not set and all trials are run
# time_limit = None
time_limit = 360

# if fast is set to True, the hyperparameter tuning will use only one fold
# however, the model will be trained on all folds in the end
# to generate OOF predictions and test predictions
# default is False
# fast = False
fast = False

Python API

To train a new model, you can run:

from autoxgb import AutoXGB


# required parameters:
train_filename = "data_samples/binary_classification.csv"
output = "output"

# optional parameters
test_filename = None
task = None
idx = None
targets = ["income"]
features = None
categorical_features = None
use_gpu = True
num_folds = 5
seed = 42
num_trials = 100
time_limit = 360
fast = False

# Now its time to train the model!
axgb = AutoXGB(
    train_filename=train_filename,
    output=output,
    test_filename=test_filename,
    task=task,
    idx=idx,
    targets=targets,
    features=features,
    categorical_features=categorical_features,
    use_gpu=use_gpu,
    num_folds=num_folds,
    seed=seed,
    num_trials=num_trials,
    time_limit=time_limit,
    fast=fast,
)
axgb.train()

CLI

Train the model using the autoxgb train command. The parameters are same as above.

autoxgb train \
 --train_filename datasets/30train.csv \
 --output outputs/30days \
 --test_filename datasets/30test.csv \
 --use_gpu

You can also serve the trained model using the autoxgb serve command.

autoxgb serve --model_path outputs/mll --host 0.0.0.0 --debug

To know more about a command, run:

`autoxgb <command> --help` 
autoxgb train --help


usage: autoxgb <command> [<args>] train [-h] --train_filename TRAIN_FILENAME [--test_filename TEST_FILENAME] --output
                                        OUTPUT [--task {classification,regression}] [--idx IDX] [--targets TARGETS]
                                        [--num_folds NUM_FOLDS] [--features FEATURES] [--use_gpu] [--fast]
                                        [--seed SEED] [--time_limit TIME_LIMIT]

optional arguments:
  -h, --help            show this help message and exit
  --train_filename TRAIN_FILENAME
                        Path to training file
  --test_filename TEST_FILENAME
                        Path to test file
  --output OUTPUT       Path to output directory
  --task {classification,regression}
                        User defined task type
  --idx IDX             ID column
  --targets TARGETS     Target column(s). If there are multiple targets, separate by ';'
  --num_folds NUM_FOLDS
                        Number of folds to use
  --features FEATURES   Features to use, separated by ';'
  --use_gpu             Whether to use GPU for training
  --fast                Whether to use fast mode for tuning params. Only one fold will be used if fast mode is set
  --seed SEED           Random seed
  --time_limit TIME_LIMIT
                        Time limit for optimization
Owner
abhishek thakur
Kaggle: www.kaggle.com/abhishek
abhishek thakur
Generate music from midi files using BPE and markov model

Generate music from midi files using BPE and markov model

Aditya Khadilkar 37 Oct 24, 2022
pure-predict: Machine learning prediction in pure Python

pure-predict speeds up and slims down machine learning prediction applications. It is a foundational tool for serverless inference or small batch prediction with popular machine learning frameworks l

Ibotta 84 Dec 29, 2022
PyPOTS - A Python Toolbox for Data Mining on Partially-Observed Time Series

A python toolbox/library for data mining on partially-observed time series, supporting tasks of forecasting/imputation/classification/clustering on incomplete multivariate time series with missing va

Wenjie Du 179 Dec 31, 2022
Forecasting prices using Facebook/Meta's Prophet model

CryptoForecasting using Machine and Deep learning (Part 1) CryptoForecasting using Machine Learning The main aspect of predicting the stock-related da

1 Nov 27, 2021
Distributed scikit-learn meta-estimators in PySpark

sk-dist: Distributed scikit-learn meta-estimators in PySpark What is it? sk-dist is a Python package for machine learning built on top of scikit-learn

Ibotta 282 Dec 09, 2022
Adaptive: parallel active learning of mathematical functions

adaptive Adaptive: parallel active learning of mathematical functions. adaptive is an open-source Python library designed to make adaptive parallel fu

741 Dec 27, 2022
A Pythonic framework for threat modeling

pytm: A Pythonic framework for threat modeling Introduction Traditional threat modeling too often comes late to the party, or sometimes not at all. In

Izar Tarandach 644 Dec 20, 2022
Deepchecks is a Python package for comprehensively validating your machine learning models and data with minimal effort

Deepchecks is a Python package for comprehensively validating your machine learning models and data with minimal effort

2.3k Jan 04, 2023
hgboost - Hyperoptimized Gradient Boosting

hgboost is short for Hyperoptimized Gradient Boosting and is a python package for hyperparameter optimization for xgboost, catboost and lightboost using cross-validation, and evaluating the results o

Erdogan Taskesen 34 Jan 03, 2023
SmartSim makes it easier to use common Machine Learning (ML) libraries like PyTorch and TensorFlow

SmartSim makes it easier to use common Machine Learning (ML) libraries like PyTorch and TensorFlow, in High Performance Computing (HPC) simulations and workloads.

Fourier-Bayesian estimation of stochastic volatility models

fourier-bayesian-sv-estimation Fourier-Bayesian estimation of stochastic volatility models Code used to run the numerical examples of "Bayesian Approa

15 Jun 20, 2022
ArviZ is a Python package for exploratory analysis of Bayesian models

ArviZ (pronounced "AR-vees") is a Python package for exploratory analysis of Bayesian models. Includes functions for posterior analysis, data storage, model checking, comparison and diagnostics

ArviZ 1.3k Jan 05, 2023
BASTA: The BAyesian STellar Algorithm

BASTA: BAyesian STellar Algorithm Current stable version: v1.0 Important note: BASTA is developed for Python 3.8, but Python 3.7 should work as well.

BASTA team 16 Nov 15, 2022
ml4h is a toolkit for machine learning on clinical data of all kinds including genetics, labs, imaging, clinical notes, and more

ml4h is a toolkit for machine learning on clinical data of all kinds including genetics, labs, imaging, clinical notes, and more

Broad Institute 65 Dec 20, 2022
Python package for stacking (machine learning technique)

vecstack Python package for stacking (stacked generalization) featuring lightweight functional API and fully compatible scikit-learn API Convenient wa

Igor Ivanov 671 Dec 25, 2022
Automated Machine Learning Pipeline with Feature Engineering and Hyper-Parameters Tuning

The mljar-supervised is an Automated Machine Learning Python package that works with tabular data. I

MLJAR 2.4k Jan 02, 2023
Dragonfly is an open source python library for scalable Bayesian optimisation.

Dragonfly is an open source python library for scalable Bayesian optimisation. Bayesian optimisation is used for optimising black-box functions whose

744 Jan 02, 2023
customer churn prediction prevention in telecom industry using machine learning and survival analysis

Telco Customer Churn Prediction - Plotly Dash Application Description This dash application allows you to predict telco customer churn using machine l

Benaissa Mohamed Fayçal 3 Nov 20, 2021
Programming assignments and quizzes from all courses within the Machine Learning Engineering for Production (MLOps) specialization offered by deeplearning.ai

Machine Learning Engineering for Production (MLOps) Specialization on Coursera (offered by deeplearning.ai) Programming assignments from all courses i

Aman Chadha 173 Jan 05, 2023
A Python-based application demonstrating various search algorithms, namely Depth-First Search (DFS), Breadth-First Search (BFS), and A* Search (Manhattan Distance Heuristic)

A Python-based application demonstrating various search algorithms, namely Depth-First Search (DFS), Breadth-First Search (BFS), and the A* Search (using the Manhattan Distance Heuristic)

17 Aug 14, 2022