hboard

中文

This project provides a web based visualization tool for experiment information. In general, Hypernets writes the experiment events into a text file. This tool could monitor the file and make these events visible.

Installation

Install with pip

pip install hboard

Install with conda

conda install -c conda-forge hboard

Install with source code

Create a frontend project after installing the front-end library hboard-frontend:

1. Create the required software environment：

• nodejs v14.15.0+

2. Besides, this project requires the package manager yarn and the module bundler webpack. Install them by the following command:

npm install -g webpack webpack-cli yarn

Clone the project codes:

git clone https://github.com/DataCanvasIO/HyperBoard.git

Build the project:

cd hboard/js

# build frontend
yarn
yarn build
rm -rf ../hboard/assets/
cp -r build/ ../hboard/assets/

# install 
cd ..
python setup.py install

Example of hboard

The folllowing steps shows how to implement the hboard：

1. Create an empty event file to store the experiment states

touch events.txt

2. Create a web server to monitor the event file

from hboard.app import WebApp
webapp = WebApp("events.txt")
webapp.start()

The web server outputs logs as：

02-24 20:45:58 I hboard.app.py 77 - experiment visualization http server is running at: http://0.0.0.0:8888

The web can be accessed via http://localhost:8888。

3. Add detailed experiment events to the file 'event.txt' by command:

echo '{"type": "experimentStart", "payload": {"task": "binary", "datasets": [{"kind": "Train", "task": "binary", "shape": [904, 17], "memory": 123072}], "steps": [{"index": 0, "name": "data_clean", "type": "DataCleanStep", "status": "wait", "configuration": {"cv": true, "data_cleaner_args": {"nan_chars": null, "correct_object_dtype": true, "drop_constant_columns": true, "drop_label_nan_rows": true, "drop_idness_columns": true, "drop_columns": null, "reserve_columns": null, "drop_duplicated_columns": false, "reduce_mem_usage": false, "int_convert_to": "float"}, "name": "data_clean", "train_test_split_strategy": null}, "extension": {}, "start_datetime": null, "end_datetime": null}, {"index": 1, "name": "space_searching", "type": "SpaceSearchStep", "status": "wait", "configuration": {"cv": true, "name": "space_searching", "num_folds": 3, "earlyStopping": {"enable": true, "exceptedReward": null, "maxNoImprovedTrials": 10, "timeLimit": 3600, "mode": "max"}}, "extension": {}, "start_datetime": null, "end_datetime": null}, {"index": 2, "name": "final_ensemble", "type": "EnsembleStep", "status": "wait", "configuration": {"ensemble_size": 20, "name": "final_ensemble", "scorer": "make_scorer(accuracy_score)"}, "extension": {}, "start_datetime": null, "end_datetime": null}], "evaluation_metric": null, "confusion_matrix": null, "resource_usage": null, "prediction_stats": null}}' >> events.txt

Now the web initialization is finished. The Hypernets experiment writes the events by callback. The web will update the lastest experiment states automatically.

Visualization when the experiment is finished

Once hboard installed, use the command 'hboard' to visualize the finished experiment.

hboard -h
usage: hboard [-h] {server} ...

hboard command is used to visualize the experiment

positional arguments:
  {server}
    server    start webserver

optional arguments:
  -h, --help  show this help message and exit

See example below:

1. Clone the codes

git clone https://github.com/DataCanvasIO/HyperBoard.git

2. Start the web server and load the experiment data

cd HyperBoard
hboard server --event-file=hboard/hboard/tests/events_example.json

View the experiment dashboard via http://localhost:8888.

Related project

Currently, HyperGBM has integrated this tool. The HyperGBM experiment could callback the web visualization function and display the experiment dashboard. Please refer to Experiment Visualization。