Spaces:
Building
Building
File size: 6,551 Bytes
f5f3483 |
1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60 61 62 63 64 65 66 67 68 69 70 71 72 73 74 75 76 77 78 79 80 81 82 83 84 85 86 87 88 89 90 91 92 93 94 95 96 97 98 99 100 101 102 103 104 105 106 107 108 109 110 111 112 113 114 115 116 117 118 119 120 121 122 123 124 125 126 127 128 129 130 131 132 133 134 135 136 137 138 139 140 141 142 143 144 145 146 147 148 149 150 151 152 153 154 155 156 157 158 159 160 161 162 163 164 165 166 167 168 169 170 171 172 173 174 175 176 177 178 179 180 181 182 183 184 185 186 187 188 189 190 191 192 193 194 195 196 197 198 199 200 201 202 203 204 205 206 207 208 209 210 211 212 213 214 215 216 217 218 219 220 221 222 223 224 225 226 227 |
# Copyright 2024 The etils Authors.
#
# Licensed under the Apache License, Version 2.0 (the "License");
# you may not use this file except in compliance with the License.
# You may obtain a copy of the License at
#
# http://www.apache.org/licenses/LICENSE-2.0
#
# Unless required by applicable law or agreed to in writing, software
# distributed under the License is distributed on an "AS IS" BASIS,
# WITHOUT WARRANTIES OR CONDITIONS OF ANY KIND, either express or implied.
# See the License for the specific language governing permissions and
# limitations under the License.
"""Utils for colab/jupyter.
Usage:
```python
from etils.ecolab import array_as_img
```
"""
from __future__ import annotations
import functools
import traceback
from typing import Any, Optional, Tuple
from etils import enp
from etils.epy import _internal
with _internal.check_missing_deps():
# pylint: disable=g-import-not-at-top
import IPython
import IPython.display
import mediapy as media
# pylint: enable=g-import-not-at-top
Array = Any
# Images smaller than this are displayed as text (e.g. `np.array([...`)
_MIN_IMG_SHAPE: Tuple[int, int] = (10, 10)
def show(*objs, **kwargs) -> None:
"""Alias for `IPython.display.display`."""
return IPython.display.display(*objs, **kwargs)
def auto_plot_array(
*,
# If updating this, also update `_array_repr_html_inner` !!!
video_min_num_frames: int = 15,
# Images outside this range are rescalled
height: None | int | tuple[int, int] = (100, 250),
show_images_kwargs: Optional[dict[str, Any]] = None,
show_videos_kwargs: Optional[dict[str, Any]] = None,
) -> None:
"""If called, 2d/3d imgage arrays will be plotted as images in colab/jupyter.
Usage:
>>> ecolab.auto_plot_array()
>>> np.zeros((28, 28, 3)) # Displayed as image
Args:
video_min_num_frames: Video `(num_frames, h, w, c)` with less than this
number of frames will be displayed as individual images
height: `(min, max)` image height in pixels. Images smaller/larger will be
reshaped. `None` to disable. If a single number, assume `min == max`.
show_images_kwargs: Kwargs forwarded to `mediapy.show_images`
show_videos_kwargs: Kwargs forwarded to `mediapy.show_videos`
"""
ipython = IPython.get_ipython()
if ipython is None:
return # Non-notebook environement
array_repr_html_fn = functools.partial(
array_repr_html,
video_min_num_frames=video_min_num_frames,
height=height,
show_images_kwargs=show_images_kwargs,
show_videos_kwargs=show_videos_kwargs,
)
# Register the new representation fo np, tf and jax array
print('Display big np/tf/jax arrays as image for nicer IPython display')
formatter = ipython.display_formatter.formatters['text/html']
# TODO(epot): How to support lazy-imports without catching everything ?
# Try registering jax
try:
jnp = enp.lazy.jnp
except ImportError:
pass
else:
# The array type is not exposed in the public API (registering jnp.ndarray
# does not works), so dynamically extracting the type
jax_array_cls = type(jnp.zeros(shape=())) # DeviceArrayBase
formatter.for_type(jax_array_cls, array_repr_html_fn)
# Try registering TF
try:
tf = enp.lazy.tf
except ImportError:
pass
else:
formatter.for_type(tf.Tensor, array_repr_html_fn)
# Try registering Torch
try:
torch = enp.lazy.torch
except ImportError:
pass
else:
formatter.for_type(torch.Tensor, array_repr_html_fn)
# Register np
formatter.for_type(enp.lazy.np.ndarray, array_repr_html_fn)
def array_repr_html(
array: Array,
**kwargs: Any,
) -> Optional[str]:
"""Returns the HTML `<img/>` repr, or `None` if array is not an image."""
try:
return _array_repr_html_inner(array, **kwargs)
except Exception:
# IPython display silence exceptions, so display it here
traceback.print_exc()
raise
def _array_repr_html_inner(
img: Array,
*,
# If updating this, also update `auto_plot_array` !!!
video_min_num_frames: int = 15,
height: None | int | tuple[int, int] = (100, 250),
show_images_kwargs: Optional[dict[str, Any]] = None,
show_videos_kwargs: Optional[dict[str, Any]] = None,
) -> Optional[str]:
"""Display the normalized img, or `None` if the input is not an image."""
show_images_kwargs = show_images_kwargs or {}
show_videos_kwargs = show_videos_kwargs or {}
if not enp.lazy.is_array(img): # Not an array
return None
# Normalize tf.Tensor into np.array
if enp.lazy.is_tf(img) or enp.lazy.is_torch(img):
img = img.numpy()
shape = img.shape
ndim = len(shape)
# Infer the array type (image or video ?)
if ndim == 2:
img_shape = shape
num_channel = 1
elif ndim == 3:
img_shape = shape[:2]
num_channel = shape[-1]
elif ndim == 4:
img_shape = shape[1:3]
num_channel = shape[-1]
num_frames = shape[0]
else:
return None
# Filter non-images
if 0 in shape: # Empty image
return None
if _smaller_than(img_shape, _MIN_IMG_SHAPE):
return None
if num_channel not in {1, 3, 4}:
return None
show_images_kwargs = show_images_kwargs.copy()
show_videos_kwargs = show_videos_kwargs.copy()
# Resize small/large images to X pixels (otherwise, difficult to see)
if height:
if isinstance(height, int):
min_height = height
max_height = height
else:
min_height, max_height = height
del height
target_height = img_shape[0] # (h, w)
target_height = max(target_height, min_height)
target_height = min(target_height, max_height)
show_images_kwargs.setdefault('height', target_height)
show_videos_kwargs.setdefault('height', target_height)
if ndim < 4:
out = media.show_image(img, return_html=True, **show_images_kwargs)
elif num_frames < video_min_num_frames:
out = media.show_images(img, return_html=True, **show_images_kwargs)
else:
# TODO(epot): media.show_video does not support single channel video
if num_channel != 3:
return None
# Dynamically compute the frame-rate, capped at 25 FPS
fps = min(num_frames // 5, 25.0)
show_videos_kwargs.setdefault('fps', fps)
out = media.show_video(
img,
return_html=True,
**show_videos_kwargs,
)
return out
def _smaller_than(shape: tuple[int, ...], min_shape: tuple[int, ...]) -> bool:
"""Returns True if one of the dim of `shape` is smaller than `min_shape`."""
return any(dim < min_dim for dim, min_dim in zip(shape, min_shape))
|