Next Release - v0.1.1 (#281)

* bumped ark version used to 0.4.2, mibi-bin-tools version to 0.2.4, and toffy version to 0.1.1

* using ark v0.4.4

* updated environment.yml

environment.yml

@@ -3,8 +3,8 @@ dependencies:
   - python=3.8
   - pip
   - pip:
-    - git+https://github.com/angelolab/ark-analysis.git@v0.3.2
-    - git+https://github.com/angelolab/[email protected].3
+    - git+https://github.com/angelolab/ark-analysis.git@v0.4.4
+    - git+https://github.com/angelolab/[email protected].4
     - jupyter>=1.0.0,<2
     - jupyter_contrib_nbextensions>=0.5.1,<1
     - jupyterlab>=3.4.3,<4

requirements.txt

@@ -1,5 +1,5 @@
-ark-analysis @ git+https://github.com/angelolab/ark-analysis.git@v0.3.2
-mibi-bin-tools @ git+https://github.com/angelolab/[email protected].3
+ark-analysis @ git+https://github.com/angelolab/ark-analysis.git@v0.4.4
+mibi-bin-tools @ git+https://github.com/angelolab/[email protected].4
 jupyter>=1.0.0,<2
 jupyter_contrib_nbextensions>=0.5.1,<1
 jupyterlab>=3.4.3,<4

setup.py

@@ -1,7 +1,7 @@
 from os import path, pardir
 from setuptools import setup, find_packages
 
-VERSION = '0.1.0'
+VERSION = '0.1.1'
 
 PKG_FOLDER = path.abspath(path.join(__file__, pardir))
 

toffy/image_stitching.py

@@ -99,7 +99,7 @@ def stitch_images(tiff_out_dir, run_dir=None, channels=None, img_sub_folder=None
     for chan in channels:
         image_data = load_utils.load_imgs_from_tree(tiff_out_dir, img_sub_folder=img_sub_folder,
                                                     fovs=folders, channels=[chan],
-                                                    max_image_size=max_img_size, dtype='float32')
+                                                    max_image_size=max_img_size)
         stitched = data_utils.stitch_images(image_data, num_cols)
         current_img = stitched.loc['stitched_image', :, :, chan].values
         fname = os.path.join(stitched_dir, chan + "_stitched.tiff")

toffy/image_utils.py

@@ -1,11 +1,13 @@
-import skimage.io as io
 import pathlib
+from typing import Dict, Union
+
 import numpy as np
-from typing import Union
+import skimage.io as io
 
 
-def save_image(fname: Union[str, pathlib.Path], data: np.ndarray,
-               compression_level: int = 6) -> None:
+def save_image(
+    fname: Union[str, pathlib.Path], data: np.ndarray, compression_level: int = 8
+) -> None:
     """
     A thin wrapper around `skimage.io.imsave()`.
 
@@ -17,10 +19,9 @@ def save_image(fname: Union[str, pathlib.Path], data: np.ndarray,
             increases compression ratio. The range of compress is `[1,9]`. Defaults to 6.
     """
     # Compression Config:
-    plugin_args: dict[str, any] = {
-        'compress': compression_level,
+    plugin_args: Dict[str, Union[str, int, Dict]] = {
+        "compression": "zlib",
+        "compressionargs": {"level": compression_level},
     }
-    if isinstance(fname, pathlib.Path):
-        fname: str = fname.as_posix()
 
     io.imsave(fname=fname, arr=data, plugin="tifffile", check_contrast=False, **plugin_args)

toffy/image_utils_test.py

@@ -1,18 +1,21 @@
-import skimage.io as io
-import numpy as np
-from toffy import image_utils
 import os
-import pytest
 import pathlib
+from typing import Iterator
+
+import numpy as np
+import pytest
+import skimage.io as io
+
+from toffy import image_utils
 
 
 @pytest.fixture(scope="session")
-def create_img_data() -> np.ndarray:
+def create_img_data() -> Iterator[np.ndarray]:
     """
     A Fixture which creates a numpy array for tiff file compression testing.
 
     Returns:
-        np.ndarray: Returns a randomly generated (1000 x 1000) numpy array.
+        Iterator[np.ndarray]: Returns a randomly generated (1000 x 1000) numpy array.
     """
 
     # Initialize a new generator - set seed for reproducibility
@@ -24,24 +27,28 @@ def create_img_data() -> np.ndarray:
     yield data
 
 
-class TestSaveImage():
+class TestSaveImage:
     @pytest.fixture(autouse=True)
     def _setup(self, tmp_path, create_img_data):
-        self.uncompressed_fname: pathlib.Path = (tmp_path / "test_img.tiff")
-        self.compressed_fname: pathlib.Path = (tmp_path / "test_img_compressed.tiff")
+        self.uncompressed_fname: pathlib.Path = tmp_path / "test_img.tiff"
+        self.compressed_fname: pathlib.Path = tmp_path / "test_img_compressed.tiff"
         self.data: np.ndarray = create_img_data
 
         # save uncompressed image
-        io.imsave(self.uncompressed_fname.as_posix(), arr=self.data, plugin="tifffile",
-                  check_contrast=False)
+        io.imsave(
+            self.uncompressed_fname,
+            arr=self.data,
+            plugin="tifffile",
+            check_contrast=False,
+        )
 
     @pytest.mark.parametrize("compress_level", [1, 6, 9,
                                                 pytest.param(10, marks=pytest.mark.xfail)])
     def test_save_compressed_img(self, compress_level):
-
         # Fails when compression_level > 9
-        image_utils.save_image(fname=self.compressed_fname, data=self.data,
-                               compression_level=compress_level)
+        image_utils.save_image(
+            fname=self.compressed_fname, data=self.data, compression_level=compress_level
+        )
 
         # Assert that the compressed tiff file is smaller than the uncompressed tiff file
         uncompressed_tiff_file_size: int = os.path.getsize(self.uncompressed_fname)
@@ -52,7 +59,7 @@ def test_save_compressed_img(self, compress_level):
         # Assert that the values in the compressed tiff file and the uncompressed
         # tiff file are equal.
 
-        uncompressed_data: np.ndarary = io.imread(self.uncompressed_fname)
+        uncompressed_data: np.ndarray = io.imread(self.uncompressed_fname)
         compressed_data: np.ndarray = io.imread(self.compressed_fname)
 
         np.testing.assert_array_equal(compressed_data, uncompressed_data)

toffy/json_utils_test.py

@@ -104,7 +104,7 @@ def test_read_json_file():
 
         # Test bad path
         bad_path = "/neasdf1246ljea/asdfje12ua3421ndsf/asdf.json"
-        with pytest.raises(ValueError, match=r'A bad path*'):
+        with pytest.raises(FileNotFoundError, match=r'A bad path*'):
             json_utils.read_json_file(bad_path)
 
         # Read json with read_json_file function assuming file path is good
@@ -127,7 +127,7 @@ def test_write_json_file():
         bad_path = "/mf8575b20d/bgjeidu45483hdck/asdf.json"
 
         # test bad path
-        with pytest.raises(ValueError, match=r"A bad path*"):
+        with pytest.raises(FileNotFoundError, match=r"A bad path*"):
             json_utils.write_json_file(json_path=bad_path, json_object=moly_json)
 
         # Write file after file path is validated

toffy/mph_comp_test.py

@@ -27,7 +27,7 @@ def test_get_estimated_time():
     good_fov = 'fov-1-scan-1'
 
     # bad directory path should raise an error
-    with pytest.raises(ValueError):
+    with pytest.raises(FileNotFoundError):
         mph.get_estimated_time(bad_path, good_fov)
 
     # bad fov name data should raise an error
@@ -88,7 +88,7 @@ def test_combine_mph_metrics():
     bad_path = os.path.join(Path(__file__).parent, "data", "not-a-folder")
 
     # bad directory path should raise an error
-    with pytest.raises(ValueError):
+    with pytest.raises(FileNotFoundError):
         mph.combine_mph_metrics(bad_path)
 
     data1 = create_sample_mph_data(fov='fov-1', mph_value=1000, total_count=50000, time=500)
@@ -121,7 +121,7 @@ def test_visualize_mph():
         }, index=[0, 1])
 
     # bad output directory path should raise an error
-    with pytest.raises(ValueError):
+    with pytest.raises(FileNotFoundError):
         mph.visualize_mph(mph_data, bad_path, regression=False)
 
     with tempfile.TemporaryDirectory() as temp_dir:

toffy/mph_inspect_test.py

@@ -51,7 +51,7 @@ def test_visualize_intensity_data(mocker):
     mass_stop = 98.5
 
     # bad path should raise an error
-    with pytest.raises(ValueError):
+    with pytest.raises(FileNotFoundError):
         mph_inspect.visualize_intensity_data(bad_path, mass, mass_start, mass_stop)
 
     # bad fov name should raise error

toffy/normalize.py

@@ -714,7 +714,7 @@ def normalize_image_data(img_dir, norm_dir, pulse_height_dir, panel_info,
 
         # get images
         images = load_utils.load_imgs_from_tree(img_dir, fovs=[fov], channels=channels,
-                                                dtype='float32', img_sub_folder=img_sub_folder)
+                                                img_sub_folder=img_sub_folder)
 
         # normalize and save
         normalize_fov(img_data=images, norm_vals=norm_vals, norm_dir=norm_dir, fov=fov,

toffy/qc_comp.py

@@ -307,8 +307,7 @@ def compute_qc_metrics(bin_file_path, extracted_imgs_path, fov_name,
 
     # retrieve the image data from extracted tiff files
     # the image coords should be: ['fov', 'type', 'x', 'y', 'channel']
-    image_data = load_utils.load_imgs_from_tree(extracted_imgs_path, fovs=[fov_name],
-                                                dtype='float32')
+    image_data = load_utils.load_imgs_from_tree(extracted_imgs_path, fovs=[fov_name])
     image_data = format_img_data(image_data)
 
     metric_csvs = compute_qc_metrics_direct(image_data, fov_name, gaussian_blur, blur_factor)

toffy/qc_metrics_plots.py

@@ -1,4 +1,4 @@
-from ark.utils.load_utils import load_imgs_from_tree, load_imgs_from_dir
+from ark.utils.load_utils import load_imgs_from_tree
 import matplotlib.pyplot as plt
 import seaborn as sns
 import numpy as np
@@ -23,8 +23,7 @@ def call_violin_swarm_plot(plotting_df, fig_label, figsize=(20, 3), fig_dir=None
     plt.xticks(rotation=45)
     if fig_dir:
         plt.savefig(fig_dir+fig_label+"_batch_effects.png", dpi=300)
-    plt.show()
-    plt.close()
+    return ax
 
 
 def make_batch_effect_plot(data_dir, normal_tissues, exclude_channels=None,

toffy/qc_metrics_plots_test.py

@@ -1,5 +1,3 @@
-from ark.utils.load_utils import load_imgs_from_tree, load_imgs_from_dir
-from ark.utils.load_utils import load_imgs_from_tree, load_imgs_from_dir
 import matplotlib.pyplot as plt
 import seaborn as sns
 import numpy as np
@@ -22,15 +20,15 @@ def test_call_violin_swarm():
 
     # Test that file is created.
     with tempfile.TemporaryDirectory() as temp_dir:
-        qc_metrics_plots.call_violin_swarm_plot(plotting_df, fig_label="test123", figsize=(20, 3),
-                                                fig_dir=temp_dir)
+        _ = qc_metrics_plots.call_violin_swarm_plot(plotting_df, fig_label="test123",
+                                                    figsize=(20, 3), fig_dir=temp_dir)
         assert os.path.exists(temp_dir+"test123"+"_batch_effects.png")
 
     # Test that no file is created when not passing fig_dir
     with tempfile.TemporaryDirectory() as temp_dir:
         file_number_before = len(os.listdir(temp_dir))
-        qc_metrics_plots.call_violin_swarm_plot(plotting_df, fig_label="test123", figsize=(20, 3),
-                                                fig_dir=None)
+        _ = qc_metrics_plots.call_violin_swarm_plot(plotting_df, fig_label="test123",
+                                                    figsize=(20, 3), fig_dir=None)
         file_number_after = len(os.listdir(temp_dir))
         assert file_number_before == file_number_after
 

toffy/rosetta.py

@@ -11,14 +11,13 @@
 import skimage.io as io
 from scipy.ndimage import gaussian_filter
 
-from ark.utils.load_utils import load_imgs_from_tree, load_imgs_from_dir
+from ark.utils.load_utils import load_imgs_from_tree
 from ark.utils.io_utils import list_folders, validate_paths, list_files, remove_file_extensions
 from ark.utils.misc_utils import verify_same_elements, verify_in_list
 from toffy.image_utils import save_image
 
 from toffy.streak_detection import streak_correction
 from toffy.json_utils import read_json_file
-from toffy import image_stitching
 
 
 def transform_compensation_json(json_path, comp_mat_path):
@@ -111,7 +110,7 @@ def validate_inputs(raw_data_dir, comp_mat, acquired_masses, acquired_targets, i
     verify_same_elements(acquired_masses=acquired_masses, compensation_masses=all_masses)
 
     # check first FOV to make sure all channels are present
-    test_data = load_imgs_from_tree(data_dir=raw_data_dir, fovs=fovs[0:1], dtype='float32',
+    test_data = load_imgs_from_tree(data_dir=raw_data_dir, fovs=fovs[0:1],
                                     img_sub_folder=raw_data_sub_folder)
 
     verify_in_list(listed_channels=acquired_targets, image_files=test_data.channels.values)
@@ -247,7 +246,7 @@ def compensate_image_data(raw_data_dir, comp_data_dir, comp_mat_path, panel_info
         # load batch of fovs
         batch_fovs = fovs[i: i + batch_size]
         batch_data = load_imgs_from_tree(data_dir=raw_data_dir, fovs=batch_fovs,
-                                         channels=acquired_targets, dtype='float32',
+                                         channels=acquired_targets,
                                          img_sub_folder=raw_data_sub_folder)
 
         # blur data
@@ -370,7 +369,7 @@ def add_source_channel_to_tiled_image(raw_img_dir, tiled_img_dir, output_dir, so
 
     # load source images
     source_imgs = load_imgs_from_tree(raw_img_dir, channels=[source_channel],
-                                      dtype='float32', img_sub_folder=img_sub_folder,
+                                      img_sub_folder=img_sub_folder,
                                       max_image_size=max_img_size)
 
     # convert stacked images to concatenated row

toffy/rosetta_test.py

@@ -204,7 +204,7 @@ def test_compensate_image_data(output_masses, input_masses, gaus_rad, save_forma
         # make fake data for testing
         fovs, chans = test_utils.gen_fov_chan_names(num_fovs=3, num_chans=3)
         filelocs, data_xr = test_utils.create_paired_xarray_fovs(
-            data_dir, fovs, chans, img_shape=(10, 10), fills=True)
+            data_dir, fovs, chans, img_shape=(10, 10), fills=True, dtype=np.float64)
         os.makedirs(os.path.join(data_dir, 'stitched_images'))
 
         # create compensation matrix
@@ -467,7 +467,7 @@ def test_copy_image_files(mocker):
                 rosetta.copy_image_files('cohort_name', ['bad_name'], temp_dir2, temp_dir)
 
             # bad paths should raise an error
-            with pytest.raises(ValueError, match='could not be found'):
+            with pytest.raises(FileNotFoundError, match='could not be found'):
                 rosetta.copy_image_files('cohort_name', run_names, 'bad_path', temp_dir)
                 rosetta.copy_image_files('cohort_name', run_names, temp_dir2, 'bad_path')
 
@@ -502,7 +502,7 @@ def test_rescale_raw_imgs():
         img_data = load_imgs_from_tree(temp_dir)
 
         # bad extracted img path should raise an error
-        with pytest.raises(ValueError):
+        with pytest.raises(FileNotFoundError):
             rosetta.rescale_raw_imgs('bad_path')
 
         # test successful image saving
@@ -558,7 +558,7 @@ def test_generate_rosetta_test_imgs(mocker):
 
         with tempfile.TemporaryDirectory() as temp_dir:
             # bad paths should raise error
-            with pytest.raises(ValueError):
+            with pytest.raises(FileNotFoundError):
                 rosetta.generate_rosetta_test_imgs('bad_path', temp_img_dir, mults,
                                                    temp_dir, panel)
                 rosetta.generate_rosetta_test_imgs(rosetta_mat_path, 'bad_path', mults,