def test_get_image_content_from_figure(self):
# given
pyplot.plot([1, 2, 3, 4])
pyplot.ylabel("some interesting numbers")
fig = pyplot.gcf()
# expect
self.assertEqual(get_image_content(fig), self._encode_figure(fig))
def test_get_image_content_from_pil_image(self):
# given
image_array = self._random_image_array()
expected_image = Image.fromarray(image_array.astype(numpy.uint8))
# expect
self.assertEqual(
get_image_content(expected_image), self._encode_pil_image(expected_image)
)
def test_get_image_content_from_2d_grayscale_array(self):
# given
image_array = self._random_image_array(d=None)
scaled_array = image_array * 255
expected_image = Image.fromarray(scaled_array.astype(numpy.uint8))
# expect
self.assertEqual(
get_image_content(image_array), self._encode_pil_image(expected_image)
)
def test_get_image_content_from_rgb_array(self):
# given
image_array = self._random_image_array()
scaled_array = image_array * 255
expected_image = Image.fromarray(scaled_array.astype(numpy.uint8))
# expect
self.assertEqual(
get_image_content(image_array), self._encode_pil_image(expected_image)
)
# and make sure that original image's size was preserved
self.assertFalse((image_array * 255 - scaled_array).any())
def test_get_image_content_from_tensorflow_tensor(self):
import tensorflow as tf # pylint: disable=C0415
# given
# pylint: disable=E1120 # false positive
image_tensor = tf.random.uniform(shape=[200, 300, 3])
expected_array = image_tensor.numpy() * 255
expected_image = Image.fromarray(expected_array.astype(numpy.uint8))
# expect
self.assertEqual(
get_image_content(image_tensor), self._encode_pil_image(expected_image)
)
def test_get_image_content_from_torch_tensor(self):
import torch # pylint: disable=C0415
# given
image_tensor = torch.rand(200, 300, 3) # pylint: disable=no-member
expected_array = image_tensor.numpy() * 255
expected_image = Image.fromarray(expected_array.astype(numpy.uint8))
# expect
self.assertEqual(
get_image_content(image_tensor), self._encode_pil_image(expected_image)
)
# and make sure that original image's size was preserved
self.assertFalse((image_tensor.numpy() * 255 - expected_array).any())
def test_get_image_content_from_3d_grayscale_array(self):
# given
image_array = numpy.array([[[1], [0]], [[-3], [4]], [[5], [6]]])
expected_array = numpy.array([[1, 0], [-3, 4], [5, 6]]) * 255
expected_image = Image.fromarray(expected_array.astype(numpy.uint8))
# expect
stderr = io.StringIO()
with contextlib.redirect_stderr(stderr):
self.assertEqual(
get_image_content(image_array), self._encode_pil_image(expected_image)
)
self.assertEqual(
stderr.getvalue(),
"The smallest value in the array is -3 and the largest value in the array is 6."
" To be interpreted as colors correctly values in the array need to be in the [0, 1] range.\n",
)
def as_image(image) -> "File":
"""Static method for converting image objects or image-like objects to an image File value object.
This way you can upload `Matplotlib` figures, `PIL` images, `NumPy` arrays, as static images.
Args:
image: Image-like object to be converted.
Supported are `PyTorch` tensors, `TensorFlow/Keras` tensors, `NumPy` arrays, `PIL` images
and `Matplotlib` figures.
Returns:
``File``: value object with converted image
Examples:
>>> import neptune.new as neptune
>>> from neptune.new.types import File
>>> run = neptune.init()
Convert NumPy array to File value object and upload it
>>> run["train/prediction_example"].upload(File.as_image(numpy_array))
Convert PIL image to File value object and upload it
>>> pil_file = File.as_image(pil_image)
>>> run["dataset/data_sample/img1"].upload(pil_file)
You can upload PIL image without explicit conversion
>>> run["dataset/data_sample/img2"].upload(pil_image)
You may also want to check `as_image docs page`_.
.. _as_image docs page:
https://docs.neptune.ai/api-reference/field-types#.as_image
"""
content_bytes = get_image_content(image)
return File.from_content(
content_bytes if content_bytes is not None else b"",
extension="png")