pingspice.analysis.evaluate.EvalSpec(object) class documentation
Part of pingspice.analysis.evaluate
(View In Hierarchy)
I hold all the information needed to specify evaluations for a given setup.
Construct me with the following arguments:
- Integer setup ID,
- List nameList of vector names,
- List indieFlags of bools, one for each vector name, indicating if its vector is independent.
- Set logValues (possibly empty) of vector names that are to be evaluated in log space rather than linear.
You will need to run my setupX
method before using the instance. If I'm not using a callable goal, you
also need to run setWeights.
| See Also | ErrorTabulator.add. | |
| Instance Variable | ivm | An instance of IV_Manager
that manages records about the independent vectors for my setup's goal
points. |
| Method | __init__ | Undocumented |
| Method | __repr__ | Undocumented |
| Method | canInterpolate | Returns True if you can use interpolation-based SSE
calculation for my setup. |
| Method | setupX | Evaluator.setup
calls this with a reference X to a goal callable or 2-D array of
goal points and a reference q to its local task queue. |
| Method | setWeights | Call this with a dict of weights, one entry keyed by each name in my nameList. |
IV_Manager
that manages records about the independent vectors for my setup's goal
points.
Returns True if you can use interpolation-based SSE
calculation for my setup.
| See Also | IV_Manager. | |
Evaluator.setup
calls this with a reference X to a goal callable or 2-D array of
goal points and a reference q to its local task queue.
I set my X attribute to the goal object. I will set my q attribute to the queue if the goal is callable and doesn't return a deferred result. That way the goal can run in a thread without blocking.
If the callable says it returns a Deferred by having a
dr attribute set True, then I will leave the q
attribute at None and trust that the callable won't block.
If there is no goal for my setup, set the X arg to
None, although that seems kind of pointless.
Call this with a dict of weights, one entry keyed by each name in my nameList.
If there isn't an entry for a particular name, it will be given a default weight of one.
If the entry is a callable, it must accept an integer ID and scalar values of its vector and return a scalar weight. In a one-time setup operation, I will call it with my ID and each goal value of its vector, and assign a weight to that goal value for my setup.
Returns a string describing the weights.