Trying to produce a causal relationships between data sources (e.g. A -> (cause) B), I could find a lot of sources for casual inference between univariate signals (e.g. temperature and speed). Nevertheless, I have hard time to find a way to test if there is casual relationship between multidimensional signal and univariate (e.g. vibration signal and speed).
More formally, supposing that we have two sources of data, one that generates multiple values each time (like a vibration sensor, or any other) and one that generate single value each time, are there any ways to discover casual connection between them (in python)?
The examples are irrelevant, I just trying to provide some context.
I already tried Granger causality test using statsmodels.tsa.stattools package (which leaded to an error). Diving into their implementation I understand that they accept "two" column data (i.e. two 1-dimensional time series).
