Skip to content
Snippets Groups Projects

Resolve "(Learner1D) add possibility to use the direct neighbors in the loss"

Merged Resolve "(Learner1D) add possibility to use the direct neighbors in the loss"
119-add-second-order-loss-to-adaptive into master
Merged Jorn Hoofwijk requested to merge 119-add-second-order-loss-to-adaptive into master
Compare
and
Show latest version
1 file
+ 57
− 6
Compare changes
  • Side-by-side
  • Inline
adaptive/learner/learner1D.py
+ 57
− 6
@@ -15,14 +15,65 @@ from ..notebook_integration import ensure_holoviews
from ..utils import cache_latest
def annotate_with_nn_neighbors(nn_neighbors):
def use_nn_neighbors(n):
"""Decorator to specify how many neighboring intervals the loss function uses.
This decorator can wrap around a loss function to let `~adaptive.Learner1D`
know that you would like to look at the N-nearest neighboring intervals.
The loss function is then guaranteed to receive the data of at least the
nn-neighbors and a dict that tells you what the neighboring points of these
are. And the Learner1D will then make sure that the loss is updated whenever
on of the nn-neighbours changes.
Examples
--------
This is a part of the curvature loss function
>>> @use_nn_neighbors(1)
...def triangle_loss(interval, scale, data, neighbors):
... x_left, x_right = interval
... xs = [neighbors[x_left][0], x_left, x_right, neighbors[x_right][1]]
... # at the boundary, neighbours[<left boundary x>] is (None, <some other x>)
... xs = [x for x in xs if x is not None]
... if len(xs) <= 2:
... return (x_right - x_left) / scale[0]
...
... y_scale = scale[1] or 1
... ys_scaled = [data[x] / y_scale for x in xs]
... xs_scaled = [x / scale[0] for x in xs]
... N = len(xs) - 2
... pts = [(x, y) for x, y in zip(xs_scaled, ys_scaled)]
... return sum(volume(pts[i:i+3]) for i in range(N)) / N
Or you may define a loss that favours the (local) minima of a function.
>>>@use_nn_neighbors(1)
...def loss(interval, scale, data, neighbors):
... x_left, x_right = interval
... n_left = neighbors[x_left][0]
... n_right = neighbors[x_right][1]
... is_min = True
...
... if n_left is not None and data[x_left] > data[n_left]:
... is_min = False
... if n_right is not None and data[x_right] > data[n_right]:
... is_min = False
...
... loss = (x_right - x_left) / scale[0]
...
... if is_min:
... return loss * 100
... return loss
"""
def _wrapped(loss_per_interval):
loss_per_interval.nn_neighbors = nn_neighbors
loss_per_interval.nn_neighbors = n
return loss_per_interval
return _wrapped
@annotate_with_nn_neighbors(0)
@use_nn_neighbors(0)
def uniform_loss(interval, scale, data, neighbors):
"""Loss function that samples the domain uniformly.
@@ -44,7 +95,7 @@ def uniform_loss(interval, scale, data, neighbors):
return dx
@annotate_with_nn_neighbors(0)
@use_nn_neighbors(0)
def default_loss(interval, scale, data, neighbors):
"""Calculate loss on a single interval.
@@ -79,7 +130,7 @@ def _loss_of_multi_interval(xs, ys):
return sum(vol(pts[i:i+3]) for i in range(N)) / N
@annotate_with_nn_neighbors(1)
@use_nn_neighbors(1)
def triangle_loss(interval, scale, data, neighbors):
x_left, x_right = interval
xs = [neighbors[x_left][0], x_left, x_right, neighbors[x_right][1]]
@@ -95,7 +146,7 @@ def triangle_loss(interval, scale, data, neighbors):
def get_curvature_loss(area_factor=1, euclid_factor=0.02, horizontal_factor=0.02):
@annotate_with_nn_neighbors(1)
@use_nn_neighbors(1)
def curvature_loss(interval, scale, data, neighbors):
triangle_loss_ = triangle_loss(interval, scale, data, neighbors)
default_loss_ = default_loss(interval, scale, data, neighbors)