Refactoring for pre- and post-processing, new design #232
Conversation
| @@ -41,7 +41,10 @@ def fit(self, points, values): | |||
|
|
|||
| if as_np_array.ndim == 1 or (as_np_array.ndim == 2 | |||
| and as_np_array.shape[1] == 1): | |||
| self.interpolator = interp1d(as_np_array, values, axis=0) | |||
|
|
|||
| self.interpolator = interp1d(as_np_array, values, axis=0, | |||
There was a problem hiding this comment.
You might want to consider adding axis=0, bounds_error=False, **kwargs to the arguments of fit(...). That way the user can control and specify these things.
There was a problem hiding this comment.
In this class, it is quite difficult to keep that general layout, mainly because we are internally using two different methods depending on the input dimension!
|
|
||
| @property | ||
| def snapshots(self): | ||
| def parameters_matrix(self): |
There was a problem hiding this comment.
small typo, could be changed to parameter_matrix
There was a problem hiding this comment.
I would keep the plural form at the moment, just to avoid changing all the tests! I'll refactor this class a bit.
|
|
||
| @property | ||
| def space(self): | ||
| def snapshots_matrix(self): |
There was a problem hiding this comment.
typo, could be changed to snapshot_matrix
There was a problem hiding this comment.
I would keep the plural form at the moment, just to avoid changing all the tests! I'll refactor this class a bit.
ezyrb/database.py
Outdated
| """ | ||
| if all(isinstance(n, int) for n in chunks): | ||
| if sum(chunks) != len(self): | ||
| raise ValueError('chunk elements are incosistent') |
There was a problem hiding this comment.
typo, change to inconsistent
ezyrb/database.py
Outdated
|
|
||
| elif all(isinstance(n, float) for n in chunks): | ||
| if not np.isclose(sum(chunks), 1.): | ||
| raise ValueError('chunk elements are incosistent') |
ezyrb/plugin/automatic_shift.py
Outdated
| The plugin implements the atomatic "shifting" preprocessing: exploiting a | ||
| machine learning framework, it is able to detect the quantity to shift the | ||
| snapshots composing the database, such that the reduction method performs | ||
| better, dipendentely by the problem at hand. |
There was a problem hiding this comment.
typo, "depending on ..."
ezyrb/plugin/automatic_shift.py
Outdated
|
|
||
| class AutomaticShiftSnapshots(Plugin): | ||
| """ | ||
| The plugin implements the atomatic "shifting" preprocessing: exploiting a |
ezyrb/plugin/automatic_shift.py
Outdated
| >>> def shift(time): | ||
| >>> return time-0.5 | ||
| >>> pod = POD() | ||
| >>> rbf = RBF() | ||
| >>> db = Database() | ||
| >>> for param in params: | ||
| >>> space, values = wave(param) | ||
| >>> snap = Snapshot(values=values, space=space) | ||
| >>> db.add(Parameter(param), snap) | ||
| >>> rom = ROM(db, pod, rbf, plugins=[ShiftSnapshots(shift, RBF())]) | ||
| >>> rom.fit() |
There was a problem hiding this comment.
This example is for ShiftSnapshots not for AutomaticShiftSnapshots
No description provided.