Trovares xGT is the fastest Deep Analytics Platform on the market and perfect for your mission-critical applications. With performance speeds 100's of times that of current market options, you can now empower your data scientist with the best development environment to meet your needs.
There are several demo notebooks available within this repo.
You need to set up a Trovares xGT server on a platform that makes sense for you, and set up a Python environment from which the Trovares xGT client is run. It is best to set up your client environment first as some Python packages are needed in order to set up some of the server environments.
The client can run from any Python environment, including Jupyter Notebook and Jupyter Lab.
One of the easiest ways to get a Jupyter environment is to install anaconda. Once installed, open an "anaconda prompt" window and run these commands:
pip install --upgrade grpcio jupyter pandas protobuf xgt
If an AWS instance will be used for running the Trovares xGT server, then you should also do this command:
pip install --upgrade boto3
An alternative to the anaconda prompt/command-line method, is to launch a Jupyter Notebook and perform these commands in notebook cells.
The first step on a Mac is to ensure that a Python (version 3.5 or later) is installed along with a pip
package.
If you are using Homebrew as your package manager, these packages can be installed by doing this command in a Terminal
window (or your favorite alternative to a bash
shell):
$ brew install python3
The first step on a native Linux platform is to ensure that a Python (version 3.5 or later) is installed.
For those distributions that use RPM and the yum
package manager:
$ sudo yum install python3 python3-devel python3-pip
On debian-based systems with the apt
package manager:
$ sudo apt install python3 python3-dev python3-pip
After these O/S packages are installed, the Python packages can be installed:
$ pip install --upgrade boto3 grpcio jupyter pandas protobuf xgt
Note that on some systems, pip
will invoke the deprecated Python2 system.
If this is the case, then you will need to do pip3
wherever pip
is shown in commands.
Also note that you may be required to install these pip
packages in your user environment.
This is done by appending --user
at the end of a pip
command.
Running the Trovares xGT application can be done in a variety of locations.
- A quick launch of Trovares xGT stack on AWS using CloudFormation:
- Running Trovares xGT on your own AWS instances from the AWS Marketplace.
- This process requires subscribing to the Trovares xGT product.
- All AWS instances with 8 vCPUs or fewer are free for the software; AWS charges for the hardware may apply.
- See Instructions for launching on the AWS Marketplace.
- An AWS cloud instance running Trovares xGT.
- The instance is run in your own account and is set up using this cloudformation.
- Launching requires use of the boto3 Python package.
- Launch AWS EC2
- A docker daemon running on your platform (on-premises).
- Any x86 system running docker can be used.
- There are two versions available;
- The
trovares/xgt
image where the application runs as the root user. - The
trovares/nonroot-xgt
image where the application runs as thexgtd
user.
- The
- Perform the equivalent of
docker pull trovares/xgt
. - More information is available at trovares/xgt and trovares/nonroot-xgt.
- A docker daemon (docker desktop) running on your laptop.
- The docker desktop hosting environment can run on:
- Windows (with WSL2 enabled)
- Mac Intel Chips
- Native Linux
- The docker desktop hosting environment can run on: