careamics 0.1.0rc1__py3-none-any.whl → 0.1.0rc2__py3-none-any.whl

careamics/utils/torch_utils.py

@@ -4,8 +4,6 @@ Convenience functions using torch.
 These functions are used to control certain aspects and behaviours of PyTorch.
 """
 import logging
-import os
-import sys
 
 import torch
 
@@ -28,66 +26,64 @@ def get_device() -> torch.device:
     return device
 
 
-def compile_model(model: torch.nn.Module) -> torch.nn.Module:
-    """
-    Torch.compile wrapper.
-
-    Parameters
-    ----------
-    model : torch.nn.Module
-        Model.
-
-    Returns
-    -------
-    torch.nn.Module
-        Compiled model if compile is available, the model itself otherwise.
-    """
-    if hasattr(torch, "compile") and sys.version_info.minor <= 9:
-        return torch.compile(model, mode="reduce-overhead")
-    else:
-        return model
-
-
-def seed_everything(seed: int) -> None:
-    """
-    Seed all random number generators for reproducibility.
-
-    Parameters
-    ----------
-    seed : int
-        Seed.
-    """
-    import random
-
-    import numpy as np
-
-    random.seed(seed)
-    np.random.seed(seed)
-    torch.manual_seed(seed)
-    torch.cuda.manual_seed_all(seed)
-
-
-def setup_cudnn_reproducibility(
-    deterministic: bool = True, benchmark: bool = True
-) -> None:
-    """
-    Prepare CuDNN benchmark and sets it to be deterministic/non-deterministic mode.
-
-    https://pytorch.org/docs/stable/notes/randomness.html#cuda-convolution-benchmarking.
-
-    Parameters
-    ----------
-    deterministic : bool
-        Deterministic mode, if running CuDNN backend.
-    benchmark : bool
-        If True, uses CuDNN heuristics to figure out which algorithm will be most
-        performant for your model architecture and input. False may slow down training.
-    """
-    if torch.cuda.is_available():
-        if deterministic:
-            deterministic = os.environ.get("CUDNN_DETERMINISTIC", "True") == "True"
-        torch.backends.cudnn.deterministic = deterministic
-
-        if benchmark:
-            benchmark = os.environ.get("CUDNN_BENCHMARK", "True") == "True"
-        torch.backends.cudnn.benchmark = benchmark
+# def compile_model(model: torch.nn.Module) -> torch.nn.Module:
+#     """
+#     Torch.compile wrapper.
+
+#     Parameters
+#     ----------
+#     model : torch.nn.Module
+#         Model.
+
+#     Returns
+#     -------
+#     torch.nn.Module
+#         Compiled model if compile is available, the model itself otherwise.
+#     """
+#     if hasattr(torch, "compile") and sys.version_info.minor <= 9:
+#         return torch.compile(model, mode="reduce-overhead")
+#     else:
+#         return model
+
+
+# def seed_everything(seed: int) -> None:
+#     """
+#     Seed all random number generators for reproducibility.
+
+#     Parameters
+#     ----------
+#     seed : int
+#         Seed.
+#     """
+#     import random
+
+#     import numpy as np
+
+#     random.seed(seed)
+#     np.random.seed(seed)
+#     torch.manual_seed(seed)
+#     torch.cuda.manual_seed_all(seed)
+
+
+# def setup_cudnn_reproducibility(
+#     deterministic: bool = True, benchmark: bool = True
+# ) -> None:
+#     """
+#     Prepare CuDNN benchmark and sets it to be deterministic/non-deterministic mode.
+
+#     Parameters
+#     ----------
+#     deterministic : bool
+#         Deterministic mode, if running CuDNN backend.
+#     benchmark : bool
+#         If True, uses CuDNN heuristics to figure out which algorithm will be most
+#         performant for your model architecture and input. False may slow down training
+#     """
+#     if torch.cuda.is_available():
+#         if deterministic:
+#             deterministic = os.environ.get("CUDNN_DETERMINISTIC", "True") == "True"
+#         torch.backends.cudnn.deterministic = deterministic
+
+#         if benchmark:
+#             benchmark = os.environ.get("CUDNN_BENCHMARK", "True") == "True"
+#         torch.backends.cudnn.benchmark = benchmark

careamics/utils/validators.py

@@ -10,7 +10,7 @@ import numpy as np
 AXES = "STCZYX"
 
 
-def check_axes_validity(axes: str) -> bool:
+def check_axes_validity(axes: str) -> None:
     """
     Sanity check on axes.
 
@@ -26,11 +26,6 @@ def check_axes_validity(axes: str) -> bool:
     ----------
     axes : str
         Axes to validate.
-
-    Returns
-    -------
-    bool
-        True if axes are valid, False otherwise.
     """
     _axes = axes.upper()
 
@@ -56,7 +51,7 @@ def check_axes_validity(axes: str) -> bool:
     if "C" in _axes:
         raise NotImplementedError("Currently, C axis is not supported.")
 
-    # prevent S and T axes together
+    # prevent S and T axes at the same time
     if "T" in _axes and "S" in _axes:
         raise NotImplementedError(
             f"Invalid axes {axes}. Cannot contain both S and T axes."
@@ -79,26 +74,45 @@ def check_axes_validity(axes: str) -> bool:
                     f"Invalid axes {axes}. Axes must be in the order {AXES}."
                 )
 
-    return True
-
 
-def check_array_validity(array: np.ndarray, axes: str) -> None:
+def add_axes(input_array: np.ndarray, axes: str) -> np.ndarray:
     """
-    Check that the numpy array is compatible with the axes.
+    Add missing axes to the input, typically batch and channel.
+
+    This method validates the axes first. Then it inspects the input array and add
+    missing dimensions if necessary.
 
     Parameters
     ----------
-    array : np.ndarray
-        Numpy array.
+    input_array : np.ndarray
+        Input array.
     axes : str
-        Valid axes (see check_axes_validity).
+        Axes to add.
+
+    Returns
+    -------
+    np.ndarray
+        Array with new singleton axes.
     """
-    if len(array.shape) - 2 != len(axes):
-        raise ValueError(
-            f"Array has {len(array.shape)} dimensions, but axes are {len(axes)}."
-            f"Externally provided arrays must have extra dimensions for batch and"
-            f"channel to be compatible with the batchnorm layers."
-        )
+    # validate axes
+    check_axes_validity(axes)
+
+    # is 3D
+    is_3D = "Z" in axes
+
+    # number of dims
+    n_dims = 5 if is_3D else 4
+
+    # array of dim 2, 3 or 4
+    if len(input_array.shape) < n_dims:
+        if "S" not in axes and "T" not in axes:
+            input_array = input_array[np.newaxis, ...]
+
+        # still missing C dimension
+        if len(input_array.shape) < n_dims:
+            input_array = input_array[:, np.newaxis, ...]
+
+    return input_array
 
 
 def check_tiling_validity(tile_shape: List[int], overlaps: List[int]) -> None:

careamics/utils/wandb.py

@@ -10,7 +10,7 @@ from typing import Dict, Union
 import torch
 import wandb
 
-from ..config import Configuration
+from careamics.config import Configuration
 
 
 def is_notebook() -> bool:

{careamics-0.1.0rc1.dist-info → careamics-0.1.0rc2.dist-info}/METADATA

@@ -1,6 +1,6 @@
 Metadata-Version: 2.1
 Name: careamics
-Version: 0.1.0rc1
+Version: 0.1.0rc2
 Summary: Toolbox for running N2V and friends.
 Project-URL: homepage, https://careamics.github.io/
 Project-URL: repository, https://github.com/CAREamics/careamics
@@ -14,6 +14,7 @@ Classifier: Programming Language :: Python :: 3.8
 Classifier: Programming Language :: Python :: 3.9
 Classifier: Programming Language :: Python :: 3.10
 Classifier: Programming Language :: Python :: 3.11
+Classifier: Programming Language :: Python :: 3.12
 Classifier: Typing :: Typed
 Requires-Python: >=3.8
 Requires-Dist: bioimageio-core
@@ -26,6 +27,7 @@ Requires-Dist: torchvision
 Requires-Dist: zarr
 Provides-Extra: all
 Requires-Dist: careamics-portfolio; extra == 'all'
+Requires-Dist: ipython; extra == 'all'
 Requires-Dist: itkwidgets; extra == 'all'
 Requires-Dist: jupyter; extra == 'all'
 Requires-Dist: pre-commit; extra == 'all'
@@ -43,12 +45,11 @@ Requires-Dist: ipython; extra == 'notebooks'
 Requires-Dist: itkwidgets; extra == 'notebooks'
 Requires-Dist: jupyter; extra == 'notebooks'
 Requires-Dist: torchsummary; extra == 'notebooks'
+Requires-Dist: wandb; extra == 'notebooks'
 Provides-Extra: test
 Requires-Dist: pytest; extra == 'test'
 Requires-Dist: pytest-cov; extra == 'test'
 Requires-Dist: wandb; extra == 'test'
-Provides-Extra: wandb
-Requires-Dist: wandb; extra == 'wandb'
 Description-Content-Type: text/markdown
 
 <p align="center">

{careamics-0.1.0rc1.dist-info → careamics-0.1.0rc2.dist-info}/RECORD

@@ -1,46 +1,47 @@
 careamics/__init__.py,sha256=loM0NCye76AI06vlOtB0sb0B_lqza4xbV5wylrhebzI,383
-careamics/engine.py,sha256=H4eXgfW-JovqRVp82FvzMG-hOgiass2-zJpdvmx1Yt8,31652
+careamics/engine.py,sha256=EAK_EtJwGrCNUGoEPPEdFit9tpnMQILhrvg-q0LMqQU,34132
 careamics/py.typed,sha256=esB4cHc6c07uVkGtqf8at7ttEnprwRxwk8obY8Qumq4,187
-careamics/bioimage/__init__.py,sha256=YhTiikELn3IVsgwBR4GcDsFKRfxr5xGU7swJxpz1R3o,314
-careamics/bioimage/io.py,sha256=6NuM1NYGXKeT0uxfjEDAsnBHRrw_GNX2YKyR4EN-Oko,7683
+careamics/bioimage/__init__.py,sha256=hXD6_jCPkqJHOBjbsmYmGiH4_kM9_ol5KtX0ECK3rgQ,334
+careamics/bioimage/io.py,sha256=Z4lqUyX9nic1pRmLIS2ulxoOTPPf_XaVHOfq38CUdAM,5533
+careamics/bioimage/rdf.py,sha256=CvVMAZqQGS2Ga-Rm0V1AxuhUTeMGOunEI_HlnD6cxiw,3003
 careamics/bioimage/docs/Noise2Void.md,sha256=tY66itOlXuVso4tHBM7zfPocjxNuQ4rya6DY7qgs_ro,271
 careamics/bioimage/docs/__init__.py,sha256=pbN_tkgjycF2Ps1PvrMKzVX7i210LbN6qxmE6lm9yVg,63
 careamics/config/__init__.py,sha256=NiHGEY0EPcyEPXJs5zPFLQ8LpJqcIN51L51AlzeBzWI,252
 careamics/config/algorithm.py,sha256=cd4G2O3ih4sLmqU4BAdwHE1UJcjfs6GC9sgGQRsVyL0,6318
-careamics/config/config.py,sha256=nDzcowyUEFC23ybP6p7Ru-Y4sITo3lKXiRlMtWwQRsI,8181
+careamics/config/config.py,sha256=TYeTQT9fzvSqzimLRSjFFAuFivzi192lZqFKgnp-Y6g,8291
 careamics/config/config_filter.py,sha256=nUPdZvvN8hb8P8rc-S938MnOL46DRK1RkOs-3KpXAh8,1082
-careamics/config/data.py,sha256=rFgt8glDijY66_qh4RcrvSPhXT-Qx6UXEjNdP1AZPQs,4982
+careamics/config/data.py,sha256=8HIHHoEwLaapX9b1UD39vBJ5EoPUFGxmGqoTXXW3ggA,5027
 careamics/config/torch_optim.py,sha256=YyicWpULiHY1jVB_MKi7HHqV5DMjvhmGOSS1tfJ3mm8,3096
 careamics/config/training.py,sha256=E21Pp8fX_NVffeemPRA_twyzEpgtg2ak1ffwx-SeTgc,15451
 careamics/dataset/__init__.py,sha256=LomFfq1GKNkTXke2Q2t3hJ8sqwfUd1DKX63svU4eL2o,22
-careamics/dataset/dataset_utils.py,sha256=ajTUDvzrmZQnnuX-S1lc0xxKPp9BGTsZ7RYiQpz9F8w,2988
+careamics/dataset/dataset_utils.py,sha256=oo2yfeTgVONTlv4R1eey0RhsoJqT4VkEQgRaReJHzeE,2822
 careamics/dataset/extraction_strategy.py,sha256=rzoggAHhXlq3h66X5C1aYG-AKG7ik-2Ifr-ESWWd1eI,485
-careamics/dataset/in_memory_dataset.py,sha256=cmEe5Dnx__eC0YEJ3IQg8Rg3vqpaNVerg-z25svGRLk,6888
-careamics/dataset/patching.py,sha256=5MaUrX225qys4abNnm04AHPs5ofB1YH1fjG388lKA-o,15403
-careamics/dataset/prepare_dataset.py,sha256=7xOvHYm2LZMMrQHdf3yVwpNgbe_k4T38aU9K4UOD4Do,5110
-careamics/dataset/tiff_dataset.py,sha256=Sje50pqA13O_bsyeW8-z4VpeoNZe4gVjzI2nkZO7x5U,7591
+careamics/dataset/in_memory_dataset.py,sha256=-2FPfj4oNNqO4HSpDstpjzgLvBJ2wby9RCHOy-7-Rew,6905
+careamics/dataset/patching.py,sha256=vVgZ4_cxYTn4ycqjYdy3XapSJKCpfGzTR8YnndXtfD4,15344
+careamics/dataset/prepare_dataset.py,sha256=-Ln9vOYNJReD1-U9kU-0hWtscWxKFc19ZrSwGTZH-mQ,5135
+careamics/dataset/tiff_dataset.py,sha256=bMdovjRC_CIlGdYPG73CWSR7R9HBSg2jLl2p4TD2WTU,7664
 careamics/losses/__init__.py,sha256=U8d0ohvezpFiUF7cBvr8Mb0tAhzZracw5i4GN-5ErOo,94
-careamics/losses/loss_factory.py,sha256=ZAIg_u9KvEqzbleSsQqLhw_08MvvuS_w-aFdsX7X0FE,757
+careamics/losses/loss_factory.py,sha256=SAkO8rv3xlQrtd21svXYLihLLK_SZfq1r86TTERAgkM,774
 careamics/losses/losses.py,sha256=HElIXbgsP1ujsvQhebl2n3KFKERIaG64K-bolI9CGDk,751
 careamics/manipulation/__init__.py,sha256=eJ5FrR0683tkEpfjVehtPXrGdrxrboOjE92gaMt54MU,119
 careamics/manipulation/pixel_manipulation.py,sha256=XeJQwHhPDjQhhLke2t_hQh4Arf9dC2wd3qxgGdG2rB4,5056
 careamics/models/__init__.py,sha256=zZE6_dg4Xv6eZZO0b9N1x-xGUyoziSpL2gOgeXRcp14,110
 careamics/models/layers.py,sha256=4xrqr6VYFKM3Eto_ll5P_h_IXzEtHOO7GES7T95cIVo,4700
-careamics/models/model_factory.py,sha256=ROvJgls01DQsvoZdmPNSu8_UN9twtkQqnP8NskJNKXo,7368
+careamics/models/model_factory.py,sha256=d3LyDt28CAhDGcLKEFDCWi4Uvb3rIsOfRaWfP1a71-Q,7892
 careamics/models/unet.py,sha256=Xon6x717SxhutOKzSsh4dCnSLesq5GrIQL4UXJ66CTY,10053
 careamics/prediction/__init__.py,sha256=zIFx-zPhpreGM5hcV5R3rxTdfWw3wnXEMg2fbYqQNfw,183
-careamics/prediction/prediction_utils.py,sha256=eK3-Xhz77mKK7UnllZSNI4ohAO6pNThhCqNT2U3UDxk,2564
-careamics/utils/__init__.py,sha256=XUMMYeM5_ZkOY9i-H-GXmBUoHUXvyugFZMsDBVX9TNE,479
+careamics/prediction/prediction_utils.py,sha256=CcSSUWw9f9xIXh5Y6QHBQC-GGaIWv_fc913_AN1wpr0,2933
+careamics/utils/__init__.py,sha256=wFqhdFsdE39x1ToFvH4iN8qAb0sMN-u4O74VWulBsJA,419
 careamics/utils/ascii_logo.txt,sha256=lWwzJbuZLEpKw8R8LU1-KEwzYWqW9G_8IZLb7z8kYO0,942
 careamics/utils/augment.py,sha256=NltOlCgUQ-kFx5M8uUrJz7JfINDxpS-Nzx9UQCH89M8,1793
 careamics/utils/context.py,sha256=4xMOMQGkfX-Sookb1BobRVBh2v8t2y4linOluHKpdec,956
 careamics/utils/logging.py,sha256=VP6QK4GqVGZiVCOHcFwrfOY6qrvMvqP6_Wj2sZpbvfc,10321
 careamics/utils/metrics.py,sha256=X5umq3mKt1J5u75imlFMJwvwkwHErAcIQfAk0bfQdrA,3484
 careamics/utils/normalization.py,sha256=8OHnv2dsuTcNBbjmYf1zjODru8Zm7oIxHZvshPJnR6g,1129
-careamics/utils/torch_utils.py,sha256=hPD_kDFloX80_WH9rOs_uEerK4xlwRMArGnYbg-tnhk,2360
-careamics/utils/validators.py,sha256=vpwCVa5X8XklqK3MvjT5IW_cgRre-nAcDHSou8SAvAM,4706
-careamics/utils/wandb.py,sha256=iSEIHesWukpwLjnUfwhdc1SuIHGjD8Pef5xiWFHUeYo,3087
-careamics-0.1.0rc1.dist-info/METADATA,sha256=MlGWIfYFGaibrTN-6CctqKJ1LXQD1NcBtOPyNqszueQ,3057
-careamics-0.1.0rc1.dist-info/WHEEL,sha256=9QBuHhg6FNW7lppboF2vKVbCGTVzsFykgRQjjlajrhA,87
-careamics-0.1.0rc1.dist-info/licenses/LICENSE,sha256=6zdNW-k_xHRKYWUf9tDI_ZplUciFHyj0g16DYuZ2udw,1509
-careamics-0.1.0rc1.dist-info/RECORD,,
+careamics/utils/torch_utils.py,sha256=vJh9AwkHvYMWMLmICHaZwXay6CJN5rg3RnuNxZG-WxY,2348
+careamics/utils/validators.py,sha256=eiAmsJs04rfDJq_iJ7WaBbeAYCuMXEd0nnEcHdGAViQ,4965
+careamics/utils/wandb.py,sha256=2h8kM99JFDsPo9Hup_Q-3fglc2QrFN-8ZGiyjZ2BZU8,3095
+careamics-0.1.0rc2.dist-info/METADATA,sha256=fs7yIZ89pIuqh-Ejca_liuT_Xlsn89qX0XZrqZtQqec,3129
+careamics-0.1.0rc2.dist-info/WHEEL,sha256=mRYSEL3Ih6g5a_CVMIcwiF__0Ae4_gLYh01YFNwiq1k,87
+careamics-0.1.0rc2.dist-info/licenses/LICENSE,sha256=6zdNW-k_xHRKYWUf9tDI_ZplUciFHyj0g16DYuZ2udw,1509
+careamics-0.1.0rc2.dist-info/RECORD,,

{careamics-0.1.0rc1.dist-info → careamics-0.1.0rc2.dist-info}/WHEEL

@@ -1,4 +1,4 @@
 Wheel-Version: 1.0
-Generator: hatchling 1.18.0
+Generator: hatchling 1.21.0
 Root-Is-Purelib: true
 Tag: py3-none-any