One of the current bottlenecks in fitting the non-relational model is the pre-computation of the sensitivity threshold.
A solution to remedy this is to allow for the pre-computation of the sensitivity threshold outside the fit function. One can specify a file containing the pre-computed value when fitting with the data. The file can be a JSON the contains the parameters for computing the sensitivity threshold, together with the results itself.
When the file is passed to the fit function, the function must first check if the parameters used in the pre-computation are consistent with the parameters passed in the fit function. Then, simply load it and skip the computation.
This is an excellent first issue if anyone is interested in contributing! :)
One of the current bottlenecks in fitting the non-relational model is the pre-computation of the sensitivity threshold.
A solution to remedy this is to allow for the pre-computation of the sensitivity threshold outside the fit function. One can specify a file containing the pre-computed value when fitting with the data. The file can be a JSON the contains the parameters for computing the sensitivity threshold, together with the results itself.
When the file is passed to the fit function, the function must first check if the parameters used in the pre-computation are consistent with the parameters passed in the fit function. Then, simply load it and skip the computation.
This is an excellent first issue if anyone is interested in contributing! :)