marsilea 0.3.1__py3-none-any.whl → 0.3.3__py3-none-any.whl

marsilea/dataset.py

@@ -1,4 +1,5 @@
 import pandas as pd
+import numpy as np
 from platformdirs import user_cache_path
 from urllib.request import urlretrieve
 
@@ -16,7 +17,8 @@ def load_data(name, cache=True):
                 (TCGA, PanCancer Atlas)
     - 'mouse_embryo': Spatial mapping of mouse embryo at E12.5
     - 'seq_align': Sequence alignment data
-    - 'les_miserables': Characters in Les Miserables
+    - 'les_miserables': Characters in Les Misérables
+    - 'sc_multiomics': Single-cell multi-omics dataset from COVID-19 patients
 
     Parameters
     ----------
@@ -33,6 +35,8 @@ def load_data(name, cache=True):
         return _load_imdb(cache)
     elif name == "pbmc3k":
         return _load_pbmc3k(cache)
+    elif name == "sc_multiomics":
+        return _load_sc_multiomics(cache)
     elif name == "oncoprint":
         return _load_oncoprint(cache)
     elif name == "mouse_embryo":
@@ -61,7 +65,11 @@ def _cache_remote(files, cache=True):
         dest = data_dir / dfname
         # Not download when cache and exist
         if not (cache and dest.exists()):
-            urlretrieve(url, dest)
+            try:
+                urlretrieve(url, dest)
+            except Exception as e:
+                print(url)
+                raise e
 
     if len(files) == 1:
         return data_dir / download_files[0]
@@ -75,28 +83,50 @@ def _load_imdb(cache=True):
 
 
 def _load_pbmc3k(cache=True):
-    exp, pct_cells, count = _cache_remote(['pbmc3k/exp.csv',
-                                           'pbmc3k/pct_cells.csv',
-                                           'pbmc3k/count.csv'],
-                                          cache=cache)
+    exp, pct_cells, count = _cache_remote(
+        ["pbmc3k/exp.csv", "pbmc3k/pct_cells.csv", "pbmc3k/count.csv"], cache=cache
+    )
     return {
-        'exp': pd.read_csv(exp, index_col=0),
-        'pct_cells': pd.read_csv(pct_cells, index_col=0),
-        'count': pd.read_csv(count, index_col=0)
+        "exp": pd.read_csv(exp, index_col=0),
+        "pct_cells": pd.read_csv(pct_cells, index_col=0),
+        "count": pd.read_csv(count, index_col=0),
     }
 
 
+def _load_sc_multiomics(cache=True):
+    stack, interaction = _cache_remote(
+        [
+            "sc-multiomics/sc-multiomics.npz",
+            "sc-multiomics/sc-multiomics-interaction.csv",
+        ]
+    )
+
+    dataset = np.load(stack, allow_pickle=True)
+    interaction = pd.read_csv(interaction)
+
+    data = {}
+    for key in dataset.files:
+        data[key] = dataset[key]
+    data["interaction"] = interaction
+
+    return data
+
+
 def _load_oncoprint(cache=True):
     cna, mrna, methyl, clinical = _cache_remote(
-        ['oncoprint/cna.csv', 'oncoprint/mrna_exp.csv',
-         'oncoprint/methyl_exp.csv', 'oncoprint/clinical.csv'],
-        cache=cache
+        [
+            "oncoprint/cna.csv",
+            "oncoprint/mrna_exp.csv",
+            "oncoprint/methyl_exp.csv",
+            "oncoprint/clinical.csv",
+        ],
+        cache=cache,
     )
     return {
-        'cna': pd.read_csv(cna, index_col=0),
-        'mrna_exp': pd.read_csv(mrna, index_col=0),
-        'methyl_exp': pd.read_csv(methyl, index_col=0),
-        'clinical': pd.read_csv(clinical, index_col=0)
+        "cna": pd.read_csv(cna, index_col=0),
+        "mrna_exp": pd.read_csv(mrna, index_col=0),
+        "methyl_exp": pd.read_csv(methyl, index_col=0),
+        "clinical": pd.read_csv(clinical, index_col=0),
     }
 
 
@@ -117,11 +147,7 @@ def _load_cooking_oils(cache=True):
 
 def _load_les_miserables(cache=True):
     nodes, links = _cache_remote(
-        ['les-miserables/miserables_nodes.csv',
-         'les-miserables/miserables_links.csv'
-         ], cache=cache
+        ["les-miserables/miserables_nodes.csv", "les-miserables/miserables_links.csv"],
+        cache=cache,
     )
-    return {
-        'nodes': pd.read_csv(nodes),
-        'links': pd.read_csv(links)
-    }
+    return {"nodes": pd.read_csv(nodes), "links": pd.read_csv(links)}

marsilea/dendrogram.py

@@ -8,17 +8,17 @@ from typing import List, Sequence
 
 
 class _DendrogramBase:
-
     is_singleton = False
 
-    def __init__(self,
-                 data,
-                 method=None,
-                 metric=None,
-                 linkage=None,
-                 get_meta_center=None,
-                 key=None,
-                 ):
+    def __init__(
+        self,
+        data,
+        method=None,
+        metric=None,
+        linkage=None,
+        get_meta_center=None,
+        key=None,
+    ):
         self.key = key
         self.data = data
         if method is None:
@@ -29,8 +29,8 @@ class _DendrogramBase:
         if len(data) == 1:
             # TODO: the y coords are heuristic value,
             #       need a better way to handle
-            self.x_coords = np.array([[1., 1., 1., 1.]])
-            self.y_coords = np.array([[0., .75, .75, 0.]])
+            self.x_coords = np.array([[1.0, 1.0, 1.0, 1.0]])
+            self.y_coords = np.array([[0.0, 0.75, 0.75, 0.0]])
             self._reorder_index = np.array([0])
             self.is_singleton = True
         else:
@@ -40,21 +40,21 @@ class _DendrogramBase:
                 self.Z = scipy_linkage(data, method=method, metric=metric)
             self._plot_data = dendrogram(self.Z, no_plot=True)
 
-            self.x_coords = np.asarray(self._plot_data['icoord']) / 5
-            self.y_coords = np.asarray(self._plot_data['dcoord'])
-            self._reorder_index = self._plot_data['leaves']
+            self.x_coords = np.asarray(self._plot_data["icoord"]) / 5
+            self.y_coords = np.asarray(self._plot_data["dcoord"])
+            self._reorder_index = self._plot_data["leaves"]
 
             if len(self.y_coords) == 1:
-                self.y_coords = np.array([[0., .75, .75, 0.]])
+                self.y_coords = np.array([[0.0, 0.75, 0.75, 0.0]])
             else:
                 ycoords = np.unique(self.y_coords)
                 ycoords = ycoords[np.nonzero(ycoords)]
                 y_min, y_max = np.min(ycoords), np.max(ycoords)
                 interval = y_max - y_min
                 for i, j in zip(*np.nonzero(self.y_coords)):
-                    if self.y_coords[i, j] != 0.:
+                    if self.y_coords[i, j] != 0.0:
                         v = self.y_coords[i, j]
-                        self.y_coords[i, j] = (v - y_min) / interval + .2
+                        self.y_coords[i, j] = (v - y_min) / interval + 0.2
                     # self.y_coords[i, j] -= offset
             # self.y_coords[np.nonzero(self.y_coords)] - offset
 
@@ -80,7 +80,8 @@ class _DendrogramBase:
             else:
                 raise ValueError(
                     "The get_meta_center must return a numpy array with shape "
-                    "matching the number of features in the data.")
+                    "matching the number of features in the data."
+                )
         else:
             raise TypeError("The get_meta_center must be a callable function or None.")
 
@@ -107,8 +108,9 @@ class _DendrogramBase:
 
         if y_end is not None:
             if y_end < self.ylim[1]:
-                raise ValueError(f"{y_end} is lower than "
-                                 f"current ylim at {self.ylim[1]}")
+                raise ValueError(
+                    f"{y_end} is lower than " f"current ylim at {self.ylim[1]}"
+                )
             self._render_ylim = (0, y_end)
 
     @property
@@ -123,7 +125,7 @@ class _DendrogramBase:
     def root(self):
         xc = self.x_coords[-1]
         yc = self.y_coords[-1]
-        x1 = (xc[2] - xc[1]) / 2. + xc[1]
+        x1 = (xc[2] - xc[1]) / 2.0 + xc[1]
         y1 = yc[1]
         return x1, y1
 
@@ -131,11 +133,11 @@ class _DendrogramBase:
     def render_root(self):
         xc = self._render_x_coords[-1]
         yc = self._render_y_coords[-1]
-        x1 = (xc[2] - xc[1]) / 2. + xc[1]
+        x1 = (xc[2] - xc[1]) / 2.0 + xc[1]
         y1 = yc[1]
         return x1, y1
 
-    def _draw_dendrogram(self, ax, orient="top", color=".1", linewidth=.7):
+    def _draw_dendrogram(self, ax, orient="top", color=".1", linewidth=0.7):
         x_coords = self._render_x_coords
         y_coords = self._render_y_coords
         if orient in ["right", "left"]:
@@ -143,14 +145,15 @@ class _DendrogramBase:
 
         lines = LineCollection(
             [list(zip(x, y)) for x, y in zip(x_coords, y_coords)],
-            color=color, linewidth=linewidth
+            color=color,
+            linewidth=linewidth,
         )
         ax.add_collection(lines)
 
 
 class Dendrogram(_DendrogramBase):
     """A dendrogram class
-    
+
     Parameters
     ----------
 
@@ -159,26 +162,39 @@ class Dendrogram(_DendrogramBase):
         Refer to :func:`scipy.cluster.hierarchy.linkage`
     metric : str
         Refer to :func:`scipy.cluster.hierarchy.linkage`
-    
+
     """
 
-    def __init__(self,
-                 data: np.ndarray,
-                 method=None,
-                 metric=None,
-                 linkage=None,
-                 get_meta_center=None,
-                 key=None,
-                 ):
-        super().__init__(data, method=method, metric=metric, key=key,
-                         linkage=linkage, get_meta_center=get_meta_center)
+    def __init__(
+        self,
+        data: np.ndarray,
+        method=None,
+        metric=None,
+        linkage=None,
+        get_meta_center=None,
+        key=None,
+    ):
+        super().__init__(
+            data,
+            method=method,
+            metric=metric,
+            key=key,
+            linkage=linkage,
+            get_meta_center=get_meta_center,
+        )
 
     # here we left an empty **kwargs to align api with GroupDendrogram
-    def draw(self, ax, orient="top",
-             color=None, linewidth=None,
-             add_root=False, root_color=None,
-             control_ax=True,
-             **kwargs):
+    def draw(
+        self,
+        ax,
+        orient="top",
+        color=None,
+        linewidth=None,
+        add_root=False,
+        root_color=None,
+        control_ax=True,
+        **kwargs,
+    ):
         """
 
         Parameters
@@ -201,10 +217,9 @@ class Dendrogram(_DendrogramBase):
         """
         color = ".1" if color is None else color
         root_color = color if root_color is None else root_color
-        linewidth = .7 if linewidth is None else linewidth
+        linewidth = 0.7 if linewidth is None else linewidth
 
-        self._draw_dendrogram(ax, orient=orient, color=color,
-                              linewidth=linewidth)
+        self._draw_dendrogram(ax, orient=orient, color=color, linewidth=linewidth)
 
         xlim = self._render_xlim
         ylim = self._render_ylim
@@ -228,8 +243,9 @@ class Dendrogram(_DendrogramBase):
             else:
                 x2 = x1
                 y2 = ylim[1]
-            root_line = Line2D([x1, x2], [y1, y2],
-                               color=root_color, linewidth=linewidth)
+            root_line = Line2D(
+                [x1, x2], [y1, y2], color=root_color, linewidth=linewidth
+            )
             ax.add_artist(root_line)
 
 
@@ -243,20 +259,22 @@ class GroupDendrogram(_DendrogramBase):
         A list of :class:`Dendrogram`
     method : str
     metric : str
-    
+
     """
 
-    def __init__(self,
-                 dens: List[Dendrogram],
-                 method=None,
-                 metric=None,
-                 get_meta_center=None,
-                 key=None,
-                 **kwargs,
-                 ):
+    def __init__(
+        self,
+        dens: List[Dendrogram],
+        method=None,
+        metric=None,
+        get_meta_center=None,
+        key=None,
+        **kwargs,
+    ):
         data = np.vstack([d.center for d in dens])
-        super().__init__(data, method=method, metric=metric,
-                         get_meta_center=get_meta_center, key=key)
+        super().__init__(
+            data, method=method, metric=metric, get_meta_center=get_meta_center, key=key
+        )
         self.orig_dens = np.asarray(dens)
         self.dens = np.asarray(dens)[self.reorder_index]
         self.n = len(self.dens)
@@ -274,19 +292,20 @@ class GroupDendrogram(_DendrogramBase):
         self.den_ylim = ylim
         self.divider = ylim * 1.05
 
-    def draw(self,
-             ax,
-             orient="top",
-             spacing=None,
-             add_meta=True,
-             add_base=True,
-             base_colors=None,
-             meta_color=None,
-             linewidth=None,
-             divide=True,
-             divide_style="--",
-             meta_ratio=.2,
-             ):
+    def draw(
+        self,
+        ax,
+        orient="top",
+        spacing=None,
+        add_meta=True,
+        add_base=True,
+        base_colors=None,
+        meta_color=None,
+        linewidth=None,
+        divide=True,
+        divide_style="--",
+        meta_ratio=0.2,
+    ):
         """
 
         Parameters
@@ -315,7 +334,7 @@ class GroupDendrogram(_DendrogramBase):
         """
 
         meta_color = ".1" if meta_color is None else meta_color
-        linewidth = .7 if linewidth is None else linewidth
+        linewidth = 0.7 if linewidth is None else linewidth
         if base_colors is None:
             base_colors = cycle([None])
         elif is_color_like(base_colors):
@@ -332,8 +351,7 @@ class GroupDendrogram(_DendrogramBase):
 
         render_xlim = self.den_xlim / (1 - np.sum(spacing))
         skeleton = np.sort(np.unique(self.x_coords[self.y_coords == 0]))
-        ranger = [(skeleton[i], skeleton[i + 1]) \
-                  for i in range(len(skeleton) - 1)]
+        ranger = [(skeleton[i], skeleton[i + 1]) for i in range(len(skeleton) - 1)]
 
         draw_dens = self.dens if add_meta else self.orig_dens
 
@@ -399,28 +417,45 @@ class GroupDendrogram(_DendrogramBase):
 
         if add_meta:
             # Add meta dendrogram
-            self._draw_dendrogram(ax, orient=orient,
-                                  color=meta_color, linewidth=linewidth)
+            self._draw_dendrogram(
+                ax, orient=orient, color=meta_color, linewidth=linewidth
+            )
 
         if divide & add_base & add_meta:
             xmin = np.min(draw_dens[0].x_coords)
             xmax = np.max(draw_dens[-1]._render_x_coords)
             if orient in ["top", "bottom"]:
-                ax.hlines(self.divider, xmin, xmax,  # 0, xlim,
-                          linestyles=divide_style, color=meta_color,
-                          linewidth=linewidth)
+                ax.hlines(
+                    self.divider,
+                    xmin,
+                    xmax,  # 0, xlim,
+                    linestyles=divide_style,
+                    color=meta_color,
+                    linewidth=linewidth,
+                )
             else:
-                ax.vlines(self.divider, xmin, xmax,  # 0, ylim,
-                          linestyles=divide_style, color=meta_color,
-                          linewidth=linewidth)
+                ax.vlines(
+                    self.divider,
+                    xmin,
+                    xmax,  # 0, ylim,
+                    linestyles=divide_style,
+                    color=meta_color,
+                    linewidth=linewidth,
+                )
 
         if add_base:
             for den, color in zip(draw_dens, base_colors):
                 # The singleton dendrogram will only be drawn if meta is drawn
                 if not den.is_singleton or add_meta:
-                    den.draw(ax, orient=orient, add_root=add_meta, color=color,
-                             linewidth=linewidth,
-                             root_color=meta_color, control_ax=False)
+                    den.draw(
+                        ax,
+                        orient=orient,
+                        add_root=add_meta,
+                        color=color,
+                        linewidth=linewidth,
+                        root_color=meta_color,
+                        control_ax=False,
+                    )
 
         xlim = render_xlim
         # reserve room to avoid clipping of the top

marsilea/exceptions.py

@@ -1,5 +1,4 @@
 class DuplicateName(Exception):
-
     def __init__(self, name):
         self.name = name
 
@@ -8,7 +7,6 @@ class DuplicateName(Exception):
 
 
 class SplitTwice(Exception):
-
     def __init__(self, axis="col"):
         self.axis = axis
 
@@ -21,11 +19,12 @@ class SplitConflict(Exception):
 
 
 class AppendLayoutError(Exception):
-
     def __str__(self):
-        return "Append a concatenated plot is not allowed," \
-               "you can only append " \
-               "plots to a concatenated plot."
+        return (
+            "Append a concatenated plot is not allowed,"
+            "you can only append "
+            "plots to a concatenated plot."
+        )
 
 
 class DataError(Exception):

marsilea/heatmap.py

@@ -2,6 +2,7 @@ from __future__ import annotations
 
 import logging
 import numpy as np
+import pandas as pd
 
 from .base import ClusterBoard
 from .plotter import ColorMesh, SizedMesh, Colors
@@ -27,22 +28,54 @@ class Heatmap(ClusterBoard):
 
     """
 
-    def __init__(self, data: np.ndarray, vmin=None, vmax=None,
-                 cmap=None, norm=None, center=None,
-                 mask=None, alpha=None, linewidth=0, linecolor="white",
-                 annot=None, fmt=None, annot_kws=None, label=None,
-                 cbar_kws=None, name=None,
-                 width=None, height=None, cluster_data=None
-                 ):
+    def __init__(
+        self,
+        data: np.ndarray,
+        vmin=None,
+        vmax=None,
+        cmap=None,
+        norm=None,
+        center=None,
+        mask=None,
+        alpha=None,
+        linewidth=0,
+        linecolor="white",
+        annot=None,
+        fmt=None,
+        annot_kws=None,
+        label=None,
+        cbar_kws=None,
+        name=None,
+        width=None,
+        height=None,
+        cluster_data=None,
+        init_main=True,
+    ):
         if cluster_data is None:
-            cluster_data = data
-        super().__init__(cluster_data, width=width, height=height,
-                         name=name)
-        mesh = ColorMesh(data, vmin=vmin, vmax=vmax, cmap=cmap,
-                         norm=norm, center=center,
-                         mask=mask, alpha=alpha, linewidth=linewidth,
-                         linecolor=linecolor, annot=annot, fmt=fmt,
-                         annot_kws=annot_kws, label=label, cbar_kws=cbar_kws)
+            if isinstance(data, pd.DataFrame):
+                cluster_data = data.values
+            else:
+                cluster_data = data
+        super().__init__(
+            cluster_data, width=width, height=height, name=name, init_main=init_main
+        )
+        mesh = ColorMesh(
+            data,
+            vmin=vmin,
+            vmax=vmax,
+            cmap=cmap,
+            norm=norm,
+            center=center,
+            mask=mask,
+            alpha=alpha,
+            linewidth=linewidth,
+            linecolor=linecolor,
+            annot=annot,
+            fmt=fmt,
+            annot_kws=annot_kws,
+            label=label,
+            cbar_kws=cbar_kws,
+        )
         name = get_plot_name(name, "main", mesh.__class__.__name__)
         mesh.set(name=name)
         self.add_layer(mesh)
@@ -65,12 +98,29 @@ class CatHeatmap(ClusterBoard):
 
     """
 
-    def __init__(self, data, palette=None, cmap=None, mask=None,
-                 name=None, width=None, height=None, cluster_data=None,
-                 linewidth=0, linecolor="white",
-                 **kwargs):
-        mesh = Colors(data, palette=palette, cmap=cmap, mask=mask,
-                      linewidth=linewidth, linecolor=linecolor, **kwargs)
+    def __init__(
+        self,
+        data,
+        palette=None,
+        cmap=None,
+        mask=None,
+        name=None,
+        width=None,
+        height=None,
+        cluster_data=None,
+        linewidth=0,
+        linecolor="white",
+        **kwargs,
+    ):
+        mesh = Colors(
+            data,
+            palette=palette,
+            cmap=cmap,
+            mask=mask,
+            linewidth=linewidth,
+            linecolor=linecolor,
+            **kwargs,
+        )
         if cluster_data is None:
             cluster_data = mesh.cluster_data
         super().__init__(cluster_data, width=width, height=height, name=name)
@@ -96,11 +146,21 @@ class SizedHeatmap(ClusterBoard):
 
     """
 
-    def __init__(self, size, color=None, cluster_data=None,
-                 name=None, width=None, height=None,
-                 **kwargs):
+    def __init__(
+        self,
+        size,
+        color=None,
+        cluster_data=None,
+        name=None,
+        width=None,
+        height=None,
+        **kwargs,
+    ):
         if cluster_data is None:
-            cluster_data = size
+            if isinstance(size, pd.DataFrame):
+                cluster_data = size.values
+            else:
+                cluster_data = size
         super().__init__(cluster_data, width=width, height=height, name=name)
 
         mesh = SizedMesh(size=size, color=color, **kwargs)