Skip to content
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

Jump to bottom

Disable clipping in Agg resamplers. #28122

Open
wants to merge 1 commit into
base: main
Choose a base branch
from
Open

Disable clipping in Agg resamplers. #28122

Show file tree
Hide file tree
Changes from all commits
Commits
Show all changes
1 commit
Select commit
File filter

Filter by extension

Filter by extension
Conversations
Failed to load comments.
Jump to
Jump to file
Failed to load files.
Diff view
Diff view
8 changes: 8 additions & 0 deletions extern/agg24-svn/include/agg_span_image_filter_gray.h
View file
Open in desktop
Original file line number Diff line number Diff line change
Expand Up @@ -397,7 +397,9 @@ namespace agg
fg += weight * *fg_ptr;

fg >>= image_filter_shift;
#ifndef MPL_DISABLE_AGG_GRAY_CLIPPING
if(fg > color_type::full_value()) fg = color_type::full_value();
#endif

span->v = (value_type)fg;
span->a = color_type::full_value();
Expand Down Expand Up @@ -491,8 +493,10 @@ namespace agg
}

fg = color_type::downshift(fg, image_filter_shift);
#ifndef MPL_DISABLE_AGG_GRAY_CLIPPING
if(fg < 0) fg = 0;
if(fg > color_type::full_value()) fg = color_type::full_value();
#endif
span->v = (value_type)fg;
span->a = color_type::full_value();

Expand Down Expand Up @@ -593,8 +597,10 @@ namespace agg
}

fg /= total_weight;
#ifndef MPL_DISABLE_AGG_GRAY_CLIPPING
if(fg < 0) fg = 0;
if(fg > color_type::full_value()) fg = color_type::full_value();
#endif

span->v = (value_type)fg;
span->a = color_type::full_value();
Expand Down Expand Up @@ -701,8 +707,10 @@ namespace agg
}

fg /= total_weight;
#ifndef MPL_DISABLE_AGG_GRAY_CLIPPING
if(fg < 0) fg = 0;
if(fg > color_type::full_value()) fg = color_type::full_value();
#endif

span->v = (value_type)fg;
span->a = color_type::full_value();
Expand Down
101 changes: 9 additions & 92 deletions lib/matplotlib/image.py
View file
Open in desktop
Original file line number Diff line number Diff line change
Expand Up @@ -427,93 +427,21 @@ def _make_image(self, A, in_bbox, out_bbox, clip_bbox, magnification=1.0,
# input data is not going to match the size on the screen so we
# have to resample to the correct number of pixels

# TODO slice input array first
a_min = A.min()
a_max = A.max()
if a_min is np.ma.masked: # All masked; values don't matter.
a_min, a_max = np.int32(0), np.int32(1)
if A.dtype.kind == 'f': # Float dtype: scale to same dtype.
scaled_dtype = np.dtype(
np.float64 if A.dtype.itemsize > 4 else np.float32)
scaled_dtype = np.dtype("f8" if A.dtype.itemsize > 4 else "f4")
if scaled_dtype.itemsize < A.dtype.itemsize:
_api.warn_external(f"Casting input data from {A.dtype}"
f" to {scaled_dtype} for imshow.")
else: # Int dtype, likely.
# TODO slice input array first
# Scale to appropriately sized float: use float32 if the
# dynamic range is small, to limit the memory footprint.
da = a_max.astype(np.float64) - a_min.astype(np.float64)
scaled_dtype = np.float64 if da > 1e8 else np.float32

# Scale the input data to [.1, .9]. The Agg interpolators clip
# to [0, 1] internally, and we use a smaller input scale to
# identify the interpolated points that need to be flagged as
# over/under. This may introduce numeric instabilities in very
# broadly scaled data.

# Always copy, and don't allow array subtypes.
A_scaled = np.array(A, dtype=scaled_dtype)
# Clip scaled data around norm if necessary. This is necessary
# for big numbers at the edge of float64's ability to represent
# changes. Applying a norm first would be good, but ruins the
# interpolation of over numbers.
self.norm.autoscale_None(A)
dv = np.float64(self.norm.vmax) - np.float64(self.norm.vmin)
vmid = np.float64(self.norm.vmin) + dv / 2
fact = 1e7 if scaled_dtype == np.float64 else 1e4
newmin = vmid - dv * fact
if newmin < a_min:
newmin = None
else:
a_min = np.float64(newmin)
newmax = vmid + dv * fact
if newmax > a_max:
newmax = None
else:
a_max = np.float64(newmax)
if newmax is not None or newmin is not None:
np.clip(A_scaled, newmin, newmax, out=A_scaled)

# Rescale the raw data to [offset, 1-offset] so that the
# resampling code will run cleanly. Using dyadic numbers here
# could reduce the error, but would not fully eliminate it and
# breaks a number of tests (due to the slightly different
# error bouncing some pixels across a boundary in the (very
# quantized) colormapping step).
offset = .1
frac = .8
# Run vmin/vmax through the same rescaling as the raw data;
# otherwise, data values close or equal to the boundaries can
# end up on the wrong side due to floating point error.
vmin, vmax = self.norm.vmin, self.norm.vmax
if vmin is np.ma.masked:
vmin, vmax = a_min, a_max
vrange = np.array([vmin, vmax], dtype=scaled_dtype)

A_scaled -= a_min
vrange -= a_min
# .item() handles a_min/a_max being ndarray subclasses.
a_min = a_min.astype(scaled_dtype).item()
a_max = a_max.astype(scaled_dtype).item()

if a_min != a_max:
A_scaled /= ((a_max - a_min) / frac)
vrange /= ((a_max - a_min) / frac)
A_scaled += offset
vrange += offset
da = A.max().astype("f8") - A.min().astype("f8")
scaled_dtype = "f8" if da > 1e8 else "f4"

# resample the input data to the correct resolution and shape
A_resampled = _resample(self, A_scaled, out_shape, t)
del A_scaled # Make sure we don't use A_scaled anymore!
# Un-scale the resampled data to approximately the original
# range. Things that interpolated to outside the original range
# will still be outside, but possibly clipped in the case of
# higher order interpolation + drastically changing data.
A_resampled -= offset
vrange -= offset
if a_min != a_max:
A_resampled *= ((a_max - a_min) / frac)
vrange *= ((a_max - a_min) / frac)
A_resampled += a_min
vrange += a_min
A_resampled = _resample(self, A.astype(scaled_dtype), out_shape, t)

# if using NoNorm, cast back to the original datatype
if isinstance(self.norm, mcolors.NoNorm):
A_resampled = A_resampled.astype(A.dtype)
Expand All @@ -536,21 +464,10 @@ def _make_image(self, A, in_bbox, out_bbox, clip_bbox, magnification=1.0,
# Apply the pixel-by-pixel alpha values if present
alpha = self.get_alpha()
if alpha is not None and np.ndim(alpha) > 0:
out_alpha *= _resample(self, alpha, out_shape,
t, resample=True)
out_alpha *= _resample(self, alpha, out_shape, t, resample=True)
# mask and run through the norm
resampled_masked = np.ma.masked_array(A_resampled, out_mask)
# we have re-set the vmin/vmax to account for small errors
# that may have moved input values in/out of range
s_vmin, s_vmax = vrange
if isinstance(self.norm, mcolors.LogNorm) and s_vmin <= 0:
# Don't give 0 or negative values to LogNorm
s_vmin = np.finfo(scaled_dtype).eps
# Block the norm from sending an update signal during the
# temporary vmin/vmax change
with self.norm.callbacks.blocked(), \
cbook._setattr_cm(self.norm, vmin=s_vmin, vmax=s_vmax):
output = self.norm(resampled_masked)
output = self.norm(resampled_masked)
else:
if A.ndim == 2: # _interpolation_stage == 'rgba'
self.norm.autoscale_None(A)
Expand Down
Binary file modified lib/matplotlib/tests/baseline_images/test_image/imshow_masked_interpolation.png
View file
Open in desktop
Sorry, something went wrong. Reload?
Sorry, we cannot display this file.
Sorry, this file is invalid so it cannot be displayed.