Global Configuration Optionsï
SimBA has a few runtime configuration options which change the global behavior of SimBA.
These are managed by python-dotenv and are stored in the simba/assets/.env file of your python installation.
Sometimes, we may want to tweak these global settings - to unlock a few extra functionalities - or, to make sure that SimBA runs more reliably in specific hardware and operating system.
After installing, and before launching SimBA using simba, you can use the following commands:
LINUXï
export PRINT_EMOJIS=False #Turns of the use of emojis in the SimBA GUI
export UNSUPERVISED_INTERFACE=True #Enables GUI access to methods for unsupervised machine learning
export NUMBA_PRECOMPILE=True #Enable precompilation of Numba-based statistical methods. Results in slower SimBA load time but removed runtime cost associated with the first iteration run of any Numba decorated functions.
export CUML=False #Enables GUI access to methods fitting supervised machine learning models using GPU device
Windowsï
set PRINT_EMOJIS=False #Turns of the use of emojis in the SimBA GUI
set UNSUPERVISED_INTERFACE=True #Enables GUI access to methods for unsupervised machine learning
export NUMBA_PRECOMPILE=True #Enable precompilation of Numba-based statistical methods. Results in slower SimBA load time but removed runtime cost associated with the first iteration run of any Numba decorated functions.
set CUML=True #Enables GUI access to methods fitting supervised machine learning models using GPU device
Windows PowerShellï
$env:PRINT_EMOJIS="False" #Turns of the use of emojis in the SimBA GUI
$env:UNSUPERVISED_INTERFACE="True" #Enables GUI access to methods for unsupervised machine learning
$env:NUMBA_PRECOMPILE="True" #Enable precompilation of Numba-based statistical methods. Results in slower SimBA load time but removed runtime cost associated with the first iteration run of any Numba decorated functions.
$env:CUML="True" #Enables GUI access to methods fitting supervised machine learning models using GPU device