Merge pull request #183 from yfukai/filtering_algorithm_update

Filtering algorithm update

.vscode/settings.json

@@ -0,0 +1,5 @@
+{
+  "python.testing.pytestArgs": ["tests"],
+  "python.testing.unittestEnabled": false,
+  "python.testing.pytestEnabled": true
+}

LICENSE.rst

@@ -29,6 +29,6 @@ The original license for MIST (https://github.com/usnistgov/MIST/blob/27a8787/LI
 
 NIST-developed software is provided by NIST as a public service. You may use, copy and distribute copies of the software in any medium, provided that you keep intact this entire notice. You may improve, modify and create derivative works of the software or any portion of the software, and you may copy and distribute such modifications or works. Modified works should carry a notice stating that you changed the software and should note the date and nature of any such change. Please explicitly acknowledge the National Institute of Standards and Technology as the source of the software.
 
-NIST-developed software is expressly provided "AS IS." NIST MAKES NO WARRANTY OF ANY KIND, EXPRESS, IMPLIED, IN FACT OR ARISING BY OPERATION OF LAW, INCLUDING, WITHOUT LIMITATION, THE IMPLIED WARRANTY OF MERCHANTABILITY, FITNESS FOR A PARTICULAR PURPOSE, NON-INFRINGEMENT AND DATA ACCURACY. NIST NEITHER REPRESENTS NOR WARRANTS THAT THE OPERATION OF THE SOFTWARE WILL BE UNINTERRUPTED OR ERROR-FREE, OR THAT ANY DEFECTS WILL BE CORRECTED. NIST DOES NOT WARRANT OR MAKE ANY REPRESENTATIONS REGARDING THE USE OF THE SOFTWARE OR THE RESULTS THEREOF, INCLUDING BUT NOT LIMITED TO THE CORRECTNESS, ACCURACY, RELIABILITY, OR USEFULNESS OF THE SOFTWARE.
+NIST-developed software is expressly provided "AS IS." NIST MAKES NO WARRANTY OF ANY KIND, EXPRESS, IMPLIED, IN FACT OR ARISING BY OPERATION OF LAsizeY, INCLUDING, WITHOUT LIMITATION, THE IMPLIED WARRANTY OF MERCHANTABILITY, FITNESS FOR A PARTICULAR PURPOSE, NON-INFRINGEMENT AND DATA ACCURACY. NIST NEITHER REPRESENTS NOR WARRANTS THAT THE OPERATION OF THE SOFTWARE WILL BE UNINTERRUPTED OR ERROR-FREE, OR THAT ANY DEFECTS WILL BE CORRECTED. NIST DOES NOT WARRANT OR MAKE ANY REPRESENTATIONS REGARDING THE USE OF THE SOFTWARE OR THE RESULTS THEREOF, INCLUDING BUT NOT LIMITED TO THE CORRECTNESS, ACCURACY, RELIABILITY, OR USEFULNESS OF THE SOFTWARE.
 
 You are solely responsible for determining the appropriateness of using and distributing the software and you assume all risks associated with its use, including but not limited to the risks and costs of program errors, compliance with applicable laws, damage to or loss of data, programs or equipment, and the unavailability or interruption of operation. This software is not intended to be used in any situation where a failure could cause risk of injury or damage to property. The software developed by NIST employees is not subject to copyright protection within the United States.

examples/stitching_example.py

@@ -12,8 +12,8 @@
 props_file_path = path.join(script_path, "../tests/data/testimages_props.csv")
 images = np.load(image_file_path)
 props = pd.read_csv(props_file_path, index_col=0)
-rows = props["row"].to_list()
-cols = props["col"].to_list()
+rows = props["col"].to_list()
+cols = props["row"].to_list()
 
 print(images.shape)
 # must be 3-dim, with each dimension meaning (tile_index,x,y)

noxfile.py

@@ -101,7 +101,10 @@ def safety(session: Session) -> None:
     """Scan dependencies for insecure packages."""
     requirements = session.poetry.export_requirements()
     session.install("safety")
-    session.run("safety", "check", "--full-report", f"--file={requirements}")
+    session.run(
+        "safety", "check", "--full-report", f"--file={requirements}", "--ignore=44715"
+    )
+    # ignore numpy update for a while as resolved in 1.22 (2022.2.13)
 
 
 @session(python=python_versions)

pyproject.toml

@@ -1,6 +1,6 @@
 [tool.poetry]
 name = "m2stitch"
-version = "0.3.0"
+version = "0.4.0"
 description = "M2Stitch"
 authors = ["Yohsuke Fukai <[email protected]>"]
 license = "MIT"
@@ -32,6 +32,7 @@ pandas-stubs = "^1.2.0"
 networkx = "^2.5.1"
 scipy = "^1.6.3"
 tqdm = "^4.60.0"
+scikit-learn = "^1.0.2"
 
 [tool.poetry.dev-dependencies]
 pytest = "^7.0.0"

src/m2stitch/__main__.py

@@ -12,8 +12,8 @@ def main() -> None:
 
 #    testimages, props = test_image_path
 #    """It exits with a status code of zero."""
-#    rows = props["row"].to_list()
-#    cols = props["col"].to_list()
+#    rows = props["col"].to_list()
+#    cols = props["row"].to_list()
 #    result_df, _ = stitch_images(testimages, rows, cols)
 
 

src/m2stitch/_constrained_refinement.py

@@ -74,23 +74,23 @@ def refine_translations(images: NumArray, grid: pd.DataFrame, r: Float) -> pd.Da
     images : NumArray
         the tile images
     grid : pd.DataFrame
-        the dataframe for the grid position, with columns "{top|left}_{x|y}_second"
+        the dataframe for the grid position, with columns "{left|top}_{x|y}_second"
     r : Float
         the repeatability
 
     Returns
     -------
     grid : pd.DataFrame
-        the refined grid position, with columns "{top|left}_{x|y|ncc}"
+        the refined grid position, with columns "{left|top}_{x|y|ncc}"
     """
-    for direction in ["top", "left"]:
+    for direction in ["left", "top"]:
         for i2, g in tqdm(grid.iterrows(), total=len(grid)):
             i1 = g[direction]
             if pd.isna(i1):
                 continue
             image1 = images[i1]
             image2 = images[i2]
-            W, H = image1.shape
+            sizeY, sizeX = image1.shape
 
             def overlap_ncc(params):
                 x, y = params
@@ -99,21 +99,27 @@ def overlap_ncc(params):
                 return ncc(subI1, subI2)
 
             init_values = [
-                int(g[f"{direction}_x_second"]),
                 int(g[f"{direction}_y_second"]),
+                int(g[f"{direction}_x_second"]),
             ]
             limits = [
-                [max(-W + 1, init_values[0] - r), min(W - 1, init_values[0] + r)],
-                [max(-H + 1, init_values[1] - r), min(H - 1, init_values[1] + r)],
+                [
+                    max(-sizeY + 1, init_values[0] - r),
+                    min(sizeY - 1, init_values[0] + r),
+                ],
+                [
+                    max(-sizeX + 1, init_values[1] - r),
+                    min(sizeX - 1, init_values[1] + r),
+                ],
             ]
             values, ncc_value = find_local_max_integer_constrained(
                 overlap_ncc, np.array(init_values), np.array(limits)
             )
-            grid.loc[i2, f"{direction}_x"] = values[0]
-            grid.loc[i2, f"{direction}_y"] = values[1]
+            grid.loc[i2, f"{direction}_y"] = values[0]
+            grid.loc[i2, f"{direction}_x"] = values[1]
             grid.loc[i2, f"{direction}_ncc"] = ncc_value
-    for direction in ["top", "left"]:
-        for xy in ["x", "y"]:
-            key = f"{direction}_{xy}"
+    for direction in ["left", "top"]:
+        for dim in "yx":
+            key = f"{direction}_{dim}"
             grid[key] = grid[key].astype(pd.Int32Dtype())
     return grid

src/m2stitch/_global_optimization.py

@@ -12,7 +12,7 @@ def compute_maximum_spanning_tree(grid: pd.DataFrame) -> nx.Graph:
     ----------
     grid : pd.DataFrame
         the dataframe for the grid position,
-        with columns "{top|left}_{x|y|ncc|valid3}"
+        with columns "{left|top}_{x|y|ncc|valid3}"
 
     Returns
     -------
@@ -21,7 +21,7 @@ def compute_maximum_spanning_tree(grid: pd.DataFrame) -> nx.Graph:
     """
     connection_graph = nx.Graph()
     for i, g in grid.iterrows():
-        for direction in ["top", "left"]:
+        for direction in ["left", "top"]:
             if not pd.isna(g[direction]):
                 weight = g[f"{direction}_ncc"]
                 if g[f"{direction}_valid3"]:
@@ -33,8 +33,8 @@ def compute_maximum_spanning_tree(grid: pd.DataFrame) -> nx.Graph:
                     direction=direction,
                     f=i,
                     t=g[direction],
-                    x=g[f"{direction}_x"],
                     y=g[f"{direction}_y"],
+                    x=g[f"{direction}_x"],
                 )
     return nx.maximum_spanning_tree(connection_graph)
 
@@ -58,8 +58,8 @@ def compute_final_position(
     grid : pd.DataFrame
         the result dataframe for the grid position, with columns "{x|y}_pos"
     """
-    grid.loc[source_index, "x_pos"] = 0
     grid.loc[source_index, "y_pos"] = 0
+    grid.loc[source_index, "x_pos"] = 0
 
     nodes = [source_index]
     walked_nodes = []
@@ -72,20 +72,19 @@ def compute_final_position(
                     props["t"] == node
                 ) & (props["f"] == adj)
                 nodes.append(adj)
-                x_pos = grid.loc[node, "x_pos"]
                 y_pos = grid.loc[node, "y_pos"]
+                x_pos = grid.loc[node, "x_pos"]
 
                 if node == props["t"]:
-                    grid.loc[adj, "x_pos"] = x_pos + props["x"]
                     grid.loc[adj, "y_pos"] = y_pos + props["y"]
+                    grid.loc[adj, "x_pos"] = x_pos + props["x"]
                 else:
-                    grid.loc[adj, "x_pos"] = x_pos - props["x"]
                     grid.loc[adj, "y_pos"] = y_pos - props["y"]
-    assert not any(pd.isna(grid["x_pos"]))
-    assert not any(pd.isna(grid["y_pos"]))
-    grid["x_pos"] = grid["x_pos"] - grid["x_pos"].min()
-    grid["y_pos"] = grid["y_pos"] - grid["y_pos"].min()
-    grid["x_pos"] = grid["x_pos"].astype(np.int32)
-    grid["y_pos"] = grid["y_pos"].astype(np.int32)
+                    grid.loc[adj, "x_pos"] = x_pos - props["x"]
+    for dim in "yx":
+        k = f"{dim}_pos"
+        assert not any(pd.isna(grid[k]))
+        grid[k] = grid[k] - grid[k].min()
+        grid[k] = grid[k].astype(np.int32)
 
     return grid