Removing deprecated functions

docs/cpp_functions.rst

@@ -115,7 +115,7 @@ The relative distance of each pixel of the ROI.
   * :download:`GooseEYE.hpp <../include/GooseEYE/GooseEYE.hpp>`
   * :ref:`Example <theory_heightheight>`.
 
-GooseEYE::Clusters
+GooseEYE::clusters
 ------------------
 
 Get clusters.

docs/examples/W2.cpp

@@ -31,7 +31,7 @@ int main()
     prrng::pcg32 rng(0);
     rowmat += rng.normal({N * N}, 0.0, (double)(M) / (double)(N));
     colmat += rng.normal({N * N}, 0.0, (double)(M) / (double)(N));
-    auto dr = rng.random({N * N}) * 2.0 + 0.1;
+    auto dr = rng.normal({N * N}, 0.0, 2.0) + 0.1;
     r = r * dr;
 
     // generate image

docs/examples/W2c.cpp

@@ -41,20 +41,30 @@ int main()
     W = xt::where(xt::equal(I, 1), 0, W);
 
     // compute individual damage clusters and their centers
-    GooseEYE::Clusters Clusters(W);
-    auto clusters = Clusters.labels();
-    auto centers = Clusters.centers();
+    auto labels = GooseEYE::clusters(W);
+    auto names = xt::unique(labels);
+    auto cpos = GooseEYE::labels_centers(labels, names);
+    auto centers = xt::zeros_like(labels);
+    int m = static_cast<int>(labels.shape(0));
+    int n = static_cast<int>(labels.shape(1));
+    for (size_t i = 0; i < names.size(); ++i) {
+        auto row = (int)round(cpos(i, 0));
+        auto col = (int)round(cpos(i, 1));
+        row = row < 0 ? 0 : (row >= m ? m - 1 : row);
+        col = col < 0 ? 0 : (col >= n ? n - 1 : col);
+        centers(row, col) = names(i);
+    }
 
     // weighted correlation
     auto WI = GooseEYE::W2({101, 101}, W, I, W);
 
     // collapsed weighted correlation
-    auto WIc = GooseEYE::W2c({101, 101}, clusters, centers, I, W);
+    auto WIc = GooseEYE::W2c({101, 101}, labels, centers, I, W);
 
     // check against previous versions
     H5Easy::File data("W2c.h5", H5Easy::File::ReadOnly);
     MYASSERT(xt::all(xt::equal(I, H5Easy::load<decltype(I)>(data, "I"))));
-    MYASSERT(xt::all(xt::equal(clusters, H5Easy::load<decltype(clusters)>(data, "clusters"))));
+    MYASSERT(xt::all(xt::equal(labels, H5Easy::load<decltype(labels)>(data, "labels"))));
     MYASSERT(xt::all(xt::equal(centers, H5Easy::load<decltype(centers)>(data, "centers"))));
     MYASSERT(xt::all(xt::equal(W, H5Easy::load<decltype(W)>(data, "W"))));
     MYASSERT(xt::allclose(WI, H5Easy::load<decltype(WI)>(data, "WI")));

docs/examples/W2c.py

@@ -30,15 +30,21 @@
 W[img == 1] = 0
 
 # compute individual damage clusters and their centers
-Clusters = GooseEYE.Clusters(W)
-clusters = Clusters.labels()
-centers = Clusters.centers()
+labels = GooseEYE.clusters(W)
+names = np.unique(labels)[1:]
+cpos = GooseEYE.labels_centers(labels, names)
+cpos = np.rint(cpos).astype(int)
+cpos[:, 0] = np.clip(cpos[:, 0], 0, labels.shape[0] - 1)
+cpos[:, 1] = np.clip(cpos[:, 1], 0, labels.shape[1] - 1)
+index = np.ravel_multi_index(cpos.T, labels.shape)
+centers = np.zeros_like(labels)
+centers.flat[index] = names
 
 # weighted correlation
 WI = GooseEYE.W2((101, 101), W, img, fmask=W)
 
 # collapsed weighted correlation
-WIc = GooseEYE.W2c((101, 101), clusters, centers, img, fmask=W)
+WIc = GooseEYE.W2c((101, 101), labels, centers, img, fmask=W)
 # </snippet>
 
 if __name__ == "__main__":
@@ -58,7 +64,7 @@
 
         with h5py.File(root / "W2c.h5", "w") as file:
             file["I"] = img
-            file["clusters"] = clusters
+            file["labels"] = labels
             file["centers"] = centers
             file["W"] = W
             file["WI"] = WI
@@ -69,7 +75,7 @@
 
         with h5py.File(root / "W2c.h5") as file:
             assert np.all(np.equal(file["I"][...], img))
-            assert np.all(np.equal(file["clusters"][...], clusters))
+            assert np.all(np.equal(file["labels"][...], labels))
             assert np.all(np.equal(file["centers"][...], centers))
             assert np.all(np.equal(file["W"][...], W))
             assert np.allclose(file["WI"][...], WI)
@@ -90,7 +96,7 @@
             np.array([[0.0, 0.0, 0.0, 0.0], [1.0, 0.0, 0.0, 1.0]], dtype="float64")
         )
 
-        fig, axes = plt.subplots(figsize=(18, 6), nrows=1, ncols=3)
+        fig, axes = plt.subplots(figsize=(20, 6), nrows=1, ncols=3, constrained_layout=True)
 
         # ---
 
@@ -152,5 +158,5 @@
         cbar.set_ticks([-phi, 0, +phi])
         cbar.set_ticklabels([r"$-\varphi$", "0", r"$+\varphi$"])
 
-        fig.savefig(root / "W2c.svg")
+        fig.savefig(root / "W2c.svg", bbox_inches="tight")
         plt.close(fig)

docs/examples/clusters.cpp

@@ -8,10 +8,10 @@ int main()
     auto I = GooseEYE::dummy_circles({100, 100}, true);
 
     // clusters
-    auto clusters = GooseEYE::clusters(I, false);
+    auto labels = GooseEYE::clusters(I, false);
 
     // clusters, if the image is periodic
-    auto clusters_periodic = GooseEYE::clusters(I, true);
+    auto labels_periodic = GooseEYE::clusters(I, true);
 
     return 0;
 }

docs/examples/clusters.py

@@ -6,10 +6,10 @@
 img = GooseEYE.dummy_circles((500, 500), periodic=True)
 
 # clusters
-clusters = GooseEYE.clusters(img, periodic=False)
+labels = GooseEYE.clusters(img, periodic=False)
 
 # clusters, if the image is periodic
-clusters_periodic = GooseEYE.clusters(img, periodic=True)
+labels_periodic = GooseEYE.clusters(img, periodic=True)
 # </snippet>
 
 if __name__ == "__main__":
@@ -30,57 +30,51 @@
 
         with h5py.File(root / "clusters.h5", "w") as file:
             file["I"] = img
-            file["clusters"] = clusters
-            file["clusters_periodic"] = clusters_periodic
+            file["clusters"] = labels
+            file["clusters_periodic"] = labels_periodic
 
     if args.check:
         import h5py
 
         with h5py.File(root / "clusters.h5") as file:
             assert np.all(np.equal(file["I"][...], img))
-            assert np.all(np.equal(file["clusters"][...], clusters))
-            assert np.all(np.equal(file["clusters_periodic"][...], clusters_periodic))
+            assert np.all(np.equal(file["clusters"][...], labels))
+            assert np.all(np.equal(file["clusters_periodic"][...], labels_periodic))
 
     if args.plot or args.show:
         import matplotlib.pyplot as plt
         import matplotlib as mpl
-        import matplotlib.cm as cm
+        import cppcolormap as cm
         import prrng
         from mpl_toolkits.axes_grid1 import make_axes_locatable
 
-        # color-scheme: modify such that the background is white
-        # N.B. for a transparent background -> 4th column == 1.
-        cmap = cm.jet(range(256))
+        names = np.unique(labels)
+        names_periodic = np.unique(labels_periodic)
+        cmap = cm.jet(names.size)
         cmap[0, :3] = 1.0
-        cmap = mpl.colors.ListedColormap(cmap)
 
-        # reshuffle for better visualisation
-        assert np.all(np.diff(np.unique(clusters)) == 1)
-        assert np.all(np.diff(np.unique(clusters_periodic)) == 1)
-        assert np.unique(clusters).size >= np.unique(clusters_periodic).size
         rng = prrng.pcg32()
-        lab = np.unique(clusters)
-        lab = lab[lab != 0]
-        new = np.copy(lab).astype(np.int64)
-        rng.shuffle(new)
-        rename = np.array(([0] + list(lab), [0] + list(new))).T
-        clusters = GooseEYE.labels_rename(clusters, rename)
-        lmap = GooseEYE.labels_map(clusters_periodic, clusters)
-        unq, unq_idx, unq_cnt = np.unique(lmap[:, 1], return_inverse=True, return_counts=True)
-        assert np.all(np.in1d(np.unique(clusters_periodic), lmap[:, 0]))
-        assert np.all(np.equal(np.sort(lmap[:, 0]), np.unique(lmap[:, 0])))
-        assert np.all(np.equal(np.sort(lmap[:, 1]), np.unique(lmap[:, 1])))
-        clusters_periodic = GooseEYE.labels_rename(clusters_periodic, lmap)
+        names = names.astype(np.int64)
+        index = 1 + np.arange(names.size - 1)
+        rng.shuffle(index)
+        cmap = cmap[[0] + list(index), :]
+
+        lmap = GooseEYE.labels_map(labels_periodic, labels)
+        assert names_periodic.size == lmap.shape[0]
+        assert np.unique(lmap[:, 0]).size == lmap.shape[0]
+        assert np.unique(lmap[:, 1]).size == lmap.shape[0]
+        labels_periodic = GooseEYE.labels_rename(labels_periodic, lmap)
 
         try:
             plt.style.use(["goose", "goose-latex"])
         except OSError:
             pass
 
-        fig, axes = plt.subplots(figsize=(8 * 4, 6), nrows=1, ncols=4)
+        fig, axes = plt.subplots(figsize=(8 * 3, 6), nrows=1, ncols=4, constrained_layout=True)
 
         ax = axes[0]
-        im = ax.imshow(img, clim=(0, 1), cmap=mpl.colors.ListedColormap(cm.gray([0, 255])))
+        bw = mpl.colors.ListedColormap(np.array([[1, 1, 1, 1], [0, 0, 0, 1]]))
+        im = ax.imshow(img, clim=(0, 1), cmap=bw)
         ax.xaxis.set_ticks([0, 500])
         ax.yaxis.set_ticks([0, 500])
         ax.set_xlim([0, 500])
@@ -94,7 +88,7 @@
         cbar.set_ticks([0, 1])
 
         ax = axes[1]
-        im = ax.imshow(clusters, clim=(0, np.max(clusters) + 1), cmap=cmap)
+        im = ax.imshow(labels, clim=(0, names.size), cmap=mpl.colors.ListedColormap(cmap))
         ax.xaxis.set_ticks([0, 500])
         ax.yaxis.set_ticks([0, 500])
         ax.set_xlim([0, 500])
@@ -104,7 +98,7 @@
         ax.set_title(r"clusters")
 
         ax = axes[2]
-        im = ax.imshow(clusters_periodic, clim=(0, np.max(clusters_periodic) + 1), cmap=cmap)
+        im = ax.imshow(labels_periodic, clim=(0, names.size), cmap=mpl.colors.ListedColormap(cmap))
         ax.xaxis.set_ticks([0, 500])
         ax.yaxis.set_ticks([0, 500])
         ax.set_xlim([0, 500])
@@ -115,9 +109,9 @@
 
         ax = axes[3]
         im = ax.imshow(
-            np.where(clusters_periodic != clusters, clusters_periodic, 0),
-            clim=(0, np.max(clusters_periodic) + 1),
-            cmap=cmap,
+            np.where(labels_periodic != labels, labels_periodic, 0),
+            clim=(0, names.size),
+            cmap=mpl.colors.ListedColormap(cmap),
         )
         ax.xaxis.set_ticks([0, 500])
         ax.yaxis.set_ticks([0, 500])
@@ -130,5 +124,5 @@
         if args.show:
             plt.show()
         else:
-            fig.savefig(root / "clusters.svg")
+            fig.savefig(root / "clusters.svg", bbox_inches="tight")
         plt.close(fig)

docs/examples/clusters_centers.cpp

@@ -15,14 +15,14 @@ int main()
     auto I = GooseEYE::dummy_circles({500, 500}, true);
 
     // clusters
-    GooseEYE::Clusters clusters(I, false);
-    auto labels = clusters.labels();
-    auto centers = clusters.center_positions();
+    auto labels = GooseEYE::clusters(I, false);
+    auto names = xt::unique(labels);
+    auto centers = GooseEYE::labels_centers(labels, names, false);
 
     // clusters, if the image is periodic
-    GooseEYE::Clusters clusters_periodic(I, true);
-    auto labels_periodic = clusters_periodic.labels();
-    auto centers_periodic = clusters_periodic.center_positions();
+    auto labels_periodic = GooseEYE::clusters(I, true);
+    auto names_periodic = xt::unique(labels_periodic);
+    auto centers_periodic = GooseEYE::labels_centers(labels_periodic, names_periodic, true);
 
     // check against previous versions
     H5Easy::File data("clusters_centers.h5", H5Easy::File::ReadOnly);

docs/examples/clusters_centers.py

@@ -6,14 +6,14 @@
 img = GooseEYE.dummy_circles((500, 500), periodic=True)
 
 # clusters
-clusters = GooseEYE.Clusters(img, periodic=False)
-labels = clusters.labels()
-centers = clusters.center_positions()
+labels = GooseEYE.clusters(img, periodic=False)
+names = np.unique(labels)
+centers = GooseEYE.labels_centers(labels, names, periodic=False)
 
 # clusters, if the image is periodic
-clusters_periodic = GooseEYE.Clusters(img, periodic=True)
-labels_periodic = clusters_periodic.labels()
-centers_periodic = clusters_periodic.center_positions()
+labels_periodic = GooseEYE.clusters(img, periodic=True)
+names_periodic = np.unique(labels_periodic)
+centers_periodic = GooseEYE.labels_centers(labels_periodic, names_periodic, periodic=True)
 # </snippet>
 
 if __name__ == "__main__":
@@ -51,19 +51,24 @@
     if args.plot:
         import matplotlib.pyplot as plt
         import matplotlib as mpl
-        import matplotlib.cm as cm
+        import cppcolormap as cm
+        import prrng
         from mpl_toolkits.axes_grid1 import make_axes_locatable
 
-        lab = np.sort(np.unique(labels))[1:]
-        lp = np.sort(np.unique(labels_periodic))[1:]
-        centers = centers[lab, :]
-        centers_periodic = centers_periodic[lp, :]
-
-        # color-scheme: modify such that the background is white
-        # N.B. for a transparent background -> 4th column == 1.
-        cmap = cm.jet(range(256))
+        cmap = cm.jet(names.size)
         cmap[0, :3] = 1.0
-        cmap = mpl.colors.ListedColormap(cmap)
+
+        rng = prrng.pcg32()
+        names = names.astype(np.int64)
+        index = 1 + np.arange(names.size - 1)
+        rng.shuffle(index)
+        cmap = cmap[[0] + list(index), :]
+
+        lmap = GooseEYE.labels_map(labels_periodic, labels)
+        assert names_periodic.size == lmap.shape[0]
+        assert np.unique(lmap[:, 0]).size == lmap.shape[0]
+        assert np.unique(lmap[:, 1]).size == lmap.shape[0]
+        labels_periodic = GooseEYE.labels_rename(labels_periodic, lmap)
 
         try:
             plt.style.use(["goose", "goose-latex"])
@@ -73,7 +78,8 @@
         fig, axes = plt.subplots(figsize=(18, 6), nrows=1, ncols=3)
 
         ax = axes[0]
-        im = ax.imshow(img, clim=(0, 1), cmap=mpl.colors.ListedColormap(cm.gray([0, 255])))
+        bw = mpl.colors.ListedColormap(np.array([[1, 1, 1, 1], [0, 0, 0, 1]]))
+        im = ax.imshow(img, clim=(0, 1), cmap=bw)
         ax.xaxis.set_ticks([0, 500])
         ax.yaxis.set_ticks([0, 500])
         ax.set_xlim([0, 500])
@@ -87,8 +93,8 @@
         cbar.set_ticks([0, 1])
 
         ax = axes[1]
-        im = ax.imshow(labels, clim=(0, np.max(labels) + 1), cmap=cmap)
-        ax.plot(centers[:, 1], centers[:, 0], ls="none", marker="o", color="r")
+        im = ax.imshow(labels, clim=(0, names.size), cmap=mpl.colors.ListedColormap(cmap))
+        ax.plot(centers[1:, 1], centers[1:, 0], ls="none", marker="+", color="k")
         ax.xaxis.set_ticks([0, 500])
         ax.yaxis.set_ticks([0, 500])
         ax.set_xlim([0, 500])
@@ -102,15 +108,15 @@
         cbar.set_ticks([])
 
         ax = axes[2]
-        im = ax.imshow(labels_periodic, clim=(0, np.max(labels) + 1), cmap=cmap)
-        ax.plot(centers_periodic[:, 1], centers_periodic[:, 0], ls="none", marker="o", color="r")
+        im = ax.imshow(labels_periodic, clim=(0, names.size), cmap=mpl.colors.ListedColormap(cmap))
+        ax.plot(centers_periodic[:, 1], centers_periodic[:, 0], ls="none", marker="+", color="k")
         ax.xaxis.set_ticks([0, 500])
         ax.yaxis.set_ticks([0, 500])
         ax.set_xlim([0, 500])
         ax.set_ylim([0, 500])
         ax.set_xlabel(r"$x$")
         ax.set_ylabel(r"$y$")
-        ax.set_title(r"clusters (periodic)")
+        ax.set_title(r"clusters, periodic (color-matched)")
         div = make_axes_locatable(ax)
         cax = div.append_axes("right", size="5%", pad=0.1)
         cbar = plt.colorbar(im, cax=cax)

docs/examples/clusters_dilate.cpp

@@ -21,7 +21,7 @@ int main()
     I(15, 16) = 1;
 
     // clusters
-    auto C = GooseEYE::Clusters(I).labels();
+    auto C = GooseEYE::clusters(I);
 
     // dilate
     auto CD = GooseEYE::dilate(C);

docs/examples/clusters_dilate.py

@@ -12,7 +12,7 @@
 img[15, 16] = True
 
 # clusters
-C = GooseEYE.Clusters(img).labels()
+C = GooseEYE.clusters(img)
 
 # dilate
 CD = GooseEYE.dilate(C)