sva is a small shell-like program that provides tools to analyze
.train file obtained by the FANN library in ASC.
sva reads a
training file for a list of bit vectors corresponding to the values
of the bits ASC has determined are excited.
sva currently allows you to view any particular excited bit
vector state and run the Fourier transform on the values of a
particular bit over the entire trace.
This program is written in Python 2 and requires the
To do list
svashould not read FANN training files as if they were sequential bit vectors. It can read the training files and allow lookup of input-output pairs. It could also determine the error rate of prediction, etc. from those pairs.
- A separate mode should be used for a trace of excited bits from a given ASC invocation. I am not sure that ASC outputs this, however.