senoquant 1.0.0b1__py3-none-any.whl

senoquant/tabs/segmentation/stardist_onnx_utils/_csbdeep/csbdeep/utils/plot_utils.py

@@ -0,0 +1,159 @@
+from __future__ import print_function, unicode_literals, absolute_import, division
+from six.moves import range, zip, map, reduce, filter
+from six import string_types
+
+import numpy as np
+
+from .utils import normalize
+
+
+
+def plot_history(history,*keys,**kwargs):
+    """Plot (Keras) training history returned by :func:`CARE.train`."""
+    import matplotlib.pyplot as plt
+
+    logy = kwargs.pop('logy',False)
+
+    if all(( isinstance(k,string_types) for k in keys )):
+        w, keys = 1, [keys]
+    else:
+        w = len(keys)
+
+    plt.gcf()
+    for i, group in enumerate(keys):
+        plt.subplot(1,w,i+1)
+        for k in ([group] if isinstance(group,string_types) else group):
+            plt.plot(history.epoch,history.history[k],'.-',label=k,**kwargs)
+            if logy:
+                plt.gca().set_yscale('log', nonposy='clip')
+        plt.xlabel('epoch')
+        plt.legend(loc='best')
+    # plt.tight_layout()
+    plt.show()
+
+
+def plot_some(*arr, **kwargs):
+    """Quickly plot multiple images at once.
+
+    each arr has to be a list of 2D or 3D images
+
+    Example
+    =======
+
+    x = np.ones((200,200))
+    plot_some([x],[x])
+
+    x = np.ones((5,200,200))
+    plot_some(x,x,x)
+
+    """
+    title_list = kwargs.pop('title_list',None)
+    pmin = kwargs.pop('pmin',0)
+    pmax = kwargs.pop('pmax',100)
+    cmap = kwargs.pop('cmap','magma')
+    imshow_kwargs = kwargs
+    return _plot_some(arr=arr, title_list=title_list, pmin=pmin, pmax=pmax, cmap=cmap, **imshow_kwargs)
+
+def _plot_some(arr, title_list=None, pmin=0, pmax=100, cmap='magma', **imshow_kwargs):
+    """
+    plots a matrix of images
+
+    arr = [ X_1, X_2, ..., X_n]
+
+    where each X_i is a list of images
+
+    :param arr:
+    :param title_list:
+    :param pmin:
+    :param pmax:
+    :param imshow_kwargs:
+    :return:
+
+    """
+    import matplotlib.pyplot as plt
+    imshow_kwargs['cmap'] = cmap
+
+    def make_acceptable(a):
+        return np.asarray(a)
+    def color_image(a):
+        return np.stack(tuple(map(to_color,a))) if 1 < a.shape[-1] <= 3 else a
+    def max_project(a):
+        ndim_allowed = 2 + int(1 <= a.shape[-1] <= 3)
+        proj_axis = tuple(range(1, 1 + max(0, a[0].ndim - ndim_allowed)))
+        return np.max(a, axis=proj_axis)
+
+    arr = map(make_acceptable,arr)
+    arr = map(color_image,arr)
+    arr = map(max_project,arr)
+    arr = list(arr)
+
+    h = len(arr)
+    w = len(arr[0])
+    plt.gcf()
+    for i in range(h):
+        for j in range(w):
+            plt.subplot(h, w, i * w + j + 1)
+            try:
+                plt.title(title_list[i][j], fontsize=8)
+            except:
+                pass
+            img = arr[i][j]
+            if pmin!=0 or pmax!=100:
+                img = normalize(img,pmin=pmin,pmax=pmax,clip=True)
+            plt.imshow(np.squeeze(img),**imshow_kwargs)
+            plt.axis("off")
+
+
+def to_color(arr, pmin=1, pmax=99.8, gamma=1., colors=((0, 1, 0), (1, 0, 1), (0, 1, 1))):
+    """Converts a 2D or 3D stack to a colored image (maximal 3 channels).
+
+    Parameters
+    ----------
+    arr : numpy.ndarray
+        2D or 3D input data
+    pmin : float
+        lower percentile, pass -1 if no lower normalization is required
+    pmax : float
+        upper percentile, pass -1 if no upper normalization is required
+    gamma : float
+        gamma correction
+    colors : list
+        list of colors (r,g,b) for each channel of the input
+
+    Returns
+    -------
+    numpy.ndarray
+        colored image
+    """
+    if not arr.ndim in (2,3):
+        raise ValueError("only 2d or 3d arrays supported")
+
+    if arr.ndim ==2:
+        arr = arr[np.newaxis]
+
+    ind_min = np.argmin(arr.shape)
+    arr = np.moveaxis(arr, ind_min, 0).astype(np.float32)
+
+    out = np.zeros(arr.shape[1:] + (3,))
+
+    eps = 1.e-20
+    if pmin>=0:
+        mi = np.percentile(arr, pmin, axis=(1, 2), keepdims=True)
+    else:
+        mi = 0
+
+    if pmax>=0:
+        ma = np.percentile(arr, pmax, axis=(1, 2), keepdims=True)
+    else:
+        ma = 1.+eps
+
+    arr_norm = (1. * arr - mi) / (ma - mi + eps)
+
+
+    for i_stack, col_stack in enumerate(colors):
+        if i_stack >= len(arr):
+            break
+        for j, c in enumerate(col_stack):
+            out[..., j] += c * arr_norm[i_stack]
+
+    return np.clip(out, 0, 1)

senoquant/tabs/segmentation/stardist_onnx_utils/_csbdeep/csbdeep/utils/six.py

@@ -0,0 +1,18 @@
+from __future__ import absolute_import, print_function
+
+try:
+    from pathlib import Path
+    Path().expanduser()
+except (ImportError,AttributeError):
+    from pathlib2 import Path
+
+try:
+    import tempfile
+    tempfile.TemporaryDirectory
+except (ImportError,AttributeError):
+    from backports import tempfile
+
+try:
+    FileNotFoundError = FileNotFoundError
+except NameError:
+    FileNotFoundError = IOError