SDK Installation Guide
To access the full power of the TripleBlind platform, download and install the SDK. To download the SDK, click on the Download SDK link in the upper-right corner of the Documentation. This downloads a .zip file to your system.
ℹ️Before using the SDK, an Access Point must be installed for your organization. See the Access Point Installation or contact your IT department to have them do so.
ℹ️It is possible for the Access Point administrator to host the SDK on behalf of the users. To learn more about this option, refer to the Hosted Jupyter Notebook SDK section in the Access Point Administration.
Prerequisites
The SDK and all examples use Python 3 (version 3.7+ is supported on all modern operating systems). The R programming language is also supported via the Reticulate package. To ensure the designed behavior and to avoid conflicts, we highly recommend operating in a Python virtual environment.
The provided installation scripts use the Conda package management system and install Miniconda along with all required libraries. The SDK should also work with other package managers (such as pip
), but currently only conda
is officially supported.
Supported Operating Systems
The SDK has been validated on modern operating systems.
- Linux (Ubuntu 18.04+, most modern distros)
- macOS 10.12.6+
- Windows 10+ (directly or via WSL2)
Networking
All SDK users will need to have the networking permission (e.g. appropriate firewall settings) to access several domains:
- The Router web interface. Typically this is “tripleblind.app”, but could be something like “acme.tripleblind.app” if running in an evaluation environment.
- The Router API. Typically this is “dev.tripleblind.app” but could be something like “devacme.tripleblind.app” if running in an evaluation environment.
- Your own Access Point.
Install the TripleBlind Python Library
These instructions and all tutorials assume the SDK has been downloaded and unzipped into a folder named ~/tripleblind-sdk
. You can safely place the SDK anywhere on your system, just adjust the instructions accordingly.
Mac/Linux installation:
1. Open a terminal window.
2. Change to the directory with the unzipped package. If you don’t already have this:
- Run:
mkdir ~/tripleblind-sdk
- Run:
cd ~/tripleblind-sdk
- Run:
curl -L https://www.tripleblind.app/sdk --output sdk.zip
- Run:
unzip sdk.zip
3. Run: ./install.sh
4. Follow the prompts to complete installation.
Windows installation:
1. Navigate to the folder where you unzipped the package.
2. Double-click on the install.ps1
script (or right click it and choose “Run with PowerShell”).
3. Follow the prompts to complete installation.
Obtain Authentication Token
Once installed, the tripleblind.yaml file in your “tripleblind-sdk” folder will need to be replaced with your specific account credentials. To update follow the instructions below.
1. Visit 🔗My Settings
2. Navigate to the Authentication tab
3. Under the SDK Configuration section click Download tripleblind.yaml
4. Replace the tripleblind.yaml file in your “tripleblind-sdk” folder with the newly downloaded tripleblind.yaml file
ℹ️The authentication token expires after 6 months, and a new token must be generated to continue using the SDK without disruption. You will receive email notifications starting 2 weeks before expiration.
ℹ️When you download a new tripleblind.yaml, the active team will be reset to your default team. You can use the tb team activate TEAM_NAME_OR_ID
command to change to a different team.
SDK Configuration
When you upgrade your SDK after a new release, a new SDK folder (with new examples or updated capabilities) will replace your current folder. For this reason, it is better to use the SDK for reference only and create your own scripts in a separate folder.
Whenever using the tripleblind
library, the tripleblind.yaml file is searched for in the current directory, and if not found then upwards through the parent directory tree. In order to allow this search, you can move the tripleblind.yaml
file into your work folder or its parent.
Alternatively, you can set the TB_API_CONFIG environment variable to point to the tripleblind.yaml
file path, and it will be automatically picked up by the tripleblind
library.
Using the SDK
The TripleBlind library and dependencies are installed in a tripleblind
conda virtual environment. From a command prompt, type conda activate tripleblind
to enter this environment.
The basic layout of the SDK is:
tutorials > notebooks
A series of Jupyter notebooks which lead you through the basics of the SDK.
demos
A handful of demonstrations covering situations where the TripleBlind technology solves real-world scenarios for businesses and other organizations.
examples
Scripts illustrating each of the major features offered by TripleBlind.
Several other important files can be found at the root of the SDK folder:
tb.py
Utility to create, search, manage, download and delete assets from the command line. Type python tb.py --help
to learn more. The Mac/Linux installer will also set up an alias so you can simply use tb
to invoke this utility.
tripleblind.yaml
Configuration file holding your personal access token and other settings.
Finally, many folders hold a README.txt
file. This provides detail about the particular area and further operating instruction.
That's it, you are ready to roll! Future upgrades can be installed by simply downloading the new SDK folder and running the installation script within it.
If you are new to TripleBlind we recommend you follow the tutorials to learn the basics before further exploring the examples and demos.