Skip to content

Fix log transforms (fixes #3809) [back port to 1.4.x] #3863

New issue

Have a question about this project? Sign up for a free GitHub account to open an issue and contact its maintainers and the community.

Sign up for GitHub

By clicking “Sign up for GitHub”, you agree to our terms of service and privacy statement. We’ll occasionally send you account related emails.

Already on GitHub? Sign in to your account

Merged
merged 7 commits into from
Dec 9, 2014
Merged

Fix log transforms (fixes #3809) [back port to 1.4.x] #3863

Changes from 1 commit
Commits
Show all changes
7 commits
Select commit Hold shift + click to select a range
1c5ee87
Ensure that IdentityTransform always returns an array.
Nov 29, 2014
b5685f4
Ensure that 'points' is an array to avoid errors when indexing (compl…
Nov 29, 2014
9dcc1fa
Even with the fix for single-point transforms (77ccc8230) it's better…
maxalbert Dec 4, 2014
49d2cf1
Add test for log transform.
maxalbert Dec 4, 2014
0fa00ef
Merge upstream master.
maxalbert Dec 4, 2014
7d35c5e
Since most transformation code assumes 2d input, convert the values t…
maxalbert Dec 4, 2014
0b43f83
Support transforming scalars as some routines need this.
maxalbert Dec 4, 2014
File filter

Filter by extension

Filter by extension
Conversations
Failed to load comments.
Loading
Jump to
Jump to file
Failed to load files.
Loading
Diff view
Diff view
Prev Previous commit
Next Next commit
Since most transformation code assumes 2d input, convert the values t… 
…o a 2d array before applying any transforms and then convert the result back to the original shape.
  • Loading branch information
@maxalbert
maxalbert committed Dec 4, 2014
commit 7d35c5edfda464bc6944d355bcf943edce994089
32 changes: 29 additions & 3 deletions lib/matplotlib/transforms.py
View file
Open in desktop
Original file line number Diff line number Diff line change
Expand Up @@ -1285,8 +1285,30 @@ def transform(self, values):

Accepts a numpy array of shape (N x :attr:`input_dims`) and
returns a numpy array of shape (N x :attr:`output_dims`).

Alternatively, accepts a numpy array of length :attr:`input_dims`
and returns a numpy array of length :attr:`output_dims`.
"""
return self.transform_affine(self.transform_non_affine(values))
# Ensure that values is a 2d array (but remember whether
# we started with a 1d or 2d array).
values = np.asanyarray(values)
ndim = values.ndim
values = values.reshape((-1, self.input_dims))

# Transform the values
res = self.transform_affine(self.transform_non_affine(values))

# Convert the result back to the shape of the input values.
if ndim == 1:
return res.reshape(-1)
elif ndim == 2:
return res
else:
raise ValueError(
"Input values must have shape (N x {dims}) "
"or ({dims}).".format(dims=self.input_dims))

return res

def transform_affine(self, values):
"""
Expand All @@ -1302,6 +1324,9 @@ def transform_affine(self, values):

Accepts a numpy array of shape (N x :attr:`input_dims`) and
returns a numpy array of shape (N x :attr:`output_dims`).

Alternatively, accepts a numpy array of length :attr:`input_dims`
and returns a numpy array of length :attr:`output_dims`.
"""
return self.get_affine().transform(values)

Expand All @@ -1318,6 +1343,9 @@ def transform_non_affine(self, values):

Accepts a numpy array of shape (N x :attr:`input_dims`) and
returns a numpy array of shape (N x :attr:`output_dims`).

Alternatively, accepts a numpy array of length :attr:`input_dims`
and returns a numpy array of length :attr:`output_dims`.
"""
return values

Expand Down Expand Up @@ -2040,8 +2068,6 @@ def __repr__(self):
return "BlendedGenericTransform(%s,%s)" % (self._x, self._y)

def transform_non_affine(self, points):
points = np.asanyarray(points).reshape((-1, 2))

if self._x.is_affine and self._y.is_affine:
return points
x = self._x
Expand Down