SpikeSort
SpikeSort copied to clipboard
btel
add documentation of components
how is it better to provide the infotamtion about what is required and provided by the component without breaking numpy convention? Should this info be included in the Notes section or in some new one?
@belevtsoff It's not clear to me what exactly are you asking. What NumPy convention are you referring to?
@cpcloud I'm referring to this numpy documentation convention. We agreed to use it in the project. Every component provides some data and may also depend on other components. This information has to be montioned in the docstring of course. I'm asking where is it better to put that information (in which section, among those proposed by the convention)
Hmm, I still don't understand. What do you mean by "component" (class, function, variable, etc?) and "data" (e.g., does a function provide data? That question is vague, do you mean it returns an array; all objects in Python provide some data, that's where my confusion lies). Assuming you mean something like the former (returning an array), are you asking where the information about dependencies should be written and what the meaning of the returned array is? An example might help me understand exactly what you mean.
Components I'm talking about are the building blocks of the spike sorting workflow. They are provided by the spike_beans high-level API so that the user can connect them together to create his own workflow. Already implemented components can be found in src/spike_beans/components.py and an example setup in the examples/sorting/cluster_beans.py.
Further info: slides - there's some info about how the high-level API is arranged data structures - the data I mentioned in the previous comment