Machine Learning Time-series Platform

System dependencies:

If you are using Anaconda, create your environment with packaged dependencies pre-installed to save time:

conda create -n mltp scipy pandas matplotlib scikit-learn pytables pip

Then activate it:

source activate mltp

Dependencies are listed in requirements.txt. Install them using:

pip install --find-links=http://wheels.scikit-image.org -r requirements.txt
Install and configure Disco

Requires: Erlang. If you see errors about escript below, this is what you're missing.
```
git clone --depth 1 git://github.com/discoproject/disco.git disco
cd disco
make
cd lib && python setup.py install && cd ..
```
Launch Disco using: bin/disco nodaemon

Navigate to localhost:8989, click "configure" and ensure that an entry for "localhost" exists under "Available nodes". If not, add one with the corresponding number of cores as the "workers" value.

Install RethinkDB

For Debian / Ubuntu:

source /etc/lsb-release && echo "deb http://download.rethinkdb.com/apt $DISTRIB_CODENAME main" | sudo tee /etc/apt/sources.list.d/rethinkdb.list
wget -qO- http://download.rethinkdb.com/apt/pubkey.gpg | sudo apt-key add -
sudo apt-get update
sudo apt-get install rethinkdb

Run rethinkdb in the MLTP directory.

Install Docker

Build the required images from Docker files by running

tools/build_docker_images.sh
Install nginx

Configuration

In the file cfg.py, locate the following variables and set them to reflect your system architecture:

PROJECT_PATH

Optionally, also set

MODELS_FOLDER, UPLOAD_FOLDER, FEATURES_FOLDER, ERR_LOG_PATH

Starting the application

First, initialize the database:

PYTHONPATH="." python flask/flask_app.py --db-init

Invoke the app either in standard mode or in debug mode:

PYTHONPATH="." python flask/flask_app.py

PYTHONPATH="." python flask/flask_app.py --debug

To specify host, port, etc., execute with --help for more information.

stefanv/cesium

Machine Learning Time-series Platform

System dependencies:

Configuration

Starting the application