While using image batch mode and setting 'Spot_radius' to 0.5, I keep getting an empty spot_img array (spot_mask.shape: (1, 1), spot_img.shape: (0, 1, 3)) and the ValueError:
ERROR::: File "/home/SPROD/sprod.py", line 312, in
ERROR:::_ = extract_img_features(input_path, img_type, input_path)
ERROR::: File "/home/feature_extraction.py", line 135, in extract_img_features
ERROR:::normed=False,
ERROR::: File "/home/skimage/feature/texture.py", line 106, in greycomatrix
ERROR:::check_nD(image, 2)
ERROR::: File "/home/skimage/_shared/utils.py", line 276, in check_nD
ERROR:::raise ValueError(msg_empty_array % (arg_name))
ERROR:::ValueError
ERROR:::The parameter image cannot be an empty array.
Otherwise, setting 'Spot_radius' to bigger integer values (such as 3, 6, 14) returns (for 'Spot_radius' set to 6) spot_mask.shape:
(13, 13), spot_img.shape: (5, 13, 3), and has another error:
ERROR::: File "/home/SPROD/sprod.py", line 312, in
ERROR:::_ = extract_img_features(input_path, img_type, input_path)
ERROR::: File "/home/SPROD/sprod/feature_extraction.py", line 120, in extract_img_features
ERROR:::spot_img = np.einsum("ij,ijk->ijk", spot_mask, spot_img)
ERROR::: File "<__array_function__ internals>", line 6, in einsum
ERROR::: File "/home/numpy/core/einsumfunc.py", line 1359, in einsum
ERROR:::return c_einsum(*operands, **kwargs)
ERROR:::ValueError
ERROR:::operands could not be broadcast together with remapped shapes [original->remapped]: (13,13)->(13,13,newaxis) (5,13,3)->(5,13,3)
My 'Spot_metadata.csv' file has the shape of (1459, 1475, 3)
By the way, do you plan on adding an alternative input to the .tif image such as an RGB numpy.ndarray for example?
While using image batch mode and setting 'Spot_radius' to 0.5, I keep getting an empty spot_img array (spot_mask.shape: (1, 1), spot_img.shape: (0, 1, 3)) and the ValueError:
ERROR::: File "/home/SPROD/sprod.py", line 312, in
ERROR:::_ = extract_img_features(input_path, img_type, input_path)
ERROR::: File "/home/feature_extraction.py", line 135, in extract_img_features
ERROR:::normed=False,
ERROR::: File "/home/skimage/feature/texture.py", line 106, in greycomatrix
ERROR:::check_nD(image, 2)
ERROR::: File "/home/skimage/_shared/utils.py", line 276, in check_nD
ERROR:::raise ValueError(msg_empty_array % (arg_name))
ERROR:::ValueError
ERROR:::The parameter
image
cannot be an empty array.Otherwise, setting 'Spot_radius' to bigger integer values (such as 3, 6, 14) returns (for 'Spot_radius' set to 6) spot_mask.shape: (13, 13), spot_img.shape: (5, 13, 3), and has another error:
ERROR::: File "/home/SPROD/sprod.py", line 312, in
ERROR:::_ = extract_img_features(input_path, img_type, input_path)
ERROR::: File "/home/SPROD/sprod/feature_extraction.py", line 120, in extract_img_features
ERROR:::spot_img = np.einsum("ij,ijk->ijk", spot_mask, spot_img)
ERROR::: File "<__array_function__ internals>", line 6, in einsum
ERROR::: File "/home/numpy/core/einsumfunc.py", line 1359, in einsum
ERROR:::return c_einsum(*operands, **kwargs)
ERROR:::ValueError
ERROR:::operands could not be broadcast together with remapped shapes [original->remapped]: (13,13)->(13,13,newaxis) (5,13,3)->(5,13,3)
My 'Spot_metadata.csv' file has the shape of (1459, 1475, 3)
By the way, do you plan on adding an alternative input to the .tif image such as an RGB numpy.ndarray for example?