cellestial 0.48.4__tar.gz → 0.49.1__tar.gz

This diff represents the content of publicly available package versions that have been released to one of the supported registries. The information contained in this diff is provided for informational purposes only and reflects changes between package versions as they appear in their respective public registries.
Files changed (65) hide show
  1. {cellestial-0.48.4 → cellestial-0.49.1}/PKG-INFO +1 -1
  2. {cellestial-0.48.4 → cellestial-0.49.1}/cellestial/datasets/datasets.py +1 -1
  3. {cellestial-0.48.4 → cellestial-0.49.1}/cellestial/frames/build.py +23 -14
  4. {cellestial-0.48.4 → cellestial-0.49.1}/cellestial/frames/operations.py +29 -14
  5. {cellestial-0.48.4 → cellestial-0.49.1}/cellestial/layers/arrow.py +125 -75
  6. {cellestial-0.48.4 → cellestial-0.49.1}/cellestial/layers/bracket.py +5 -6
  7. {cellestial-0.48.4 → cellestial-0.49.1}/cellestial/layers/outline.py +1 -1
  8. {cellestial-0.48.4 → cellestial-0.49.1}/cellestial/single/core/dimensional.py +2 -2
  9. {cellestial-0.48.4 → cellestial-0.49.1}/cellestial/single/core/distribution.py +3 -209
  10. {cellestial-0.48.4 → cellestial-0.49.1}/cellestial/single/core/subdimensional.py +8 -8
  11. {cellestial-0.48.4 → cellestial-0.49.1}/cellestial/single/core/subdimensionals.py +10 -10
  12. cellestial-0.49.1/cellestial/single/core/utilities.py +221 -0
  13. {cellestial-0.48.4 → cellestial-0.49.1}/cellestial/single/differential/markers.py +2 -145
  14. cellestial-0.49.1/cellestial/single/differential/utilities.py +309 -0
  15. {cellestial-0.48.4 → cellestial-0.49.1}/cellestial/single/differential/volcano.py +1 -152
  16. {cellestial-0.48.4 → cellestial-0.49.1}/cellestial/single/heatmap/dotplot.py +2 -2
  17. {cellestial-0.48.4 → cellestial-0.49.1}/cellestial/single/heatmap/heatmap.py +25 -148
  18. {cellestial-0.48.4 → cellestial-0.49.1}/cellestial/single/heatmap/stacked_violin.py +10 -134
  19. cellestial-0.49.1/cellestial/single/heatmap/utilities.py +872 -0
  20. {cellestial-0.48.4 → cellestial-0.49.1}/cellestial/spatial/spatial.py +1 -90
  21. cellestial-0.48.4/cellestial/spatial/_components.py → cellestial-0.49.1/cellestial/spatial/utilities.py +95 -10
  22. {cellestial-0.48.4 → cellestial-0.49.1}/cellestial/util/dendrogram.py +1 -4
  23. {cellestial-0.48.4 → cellestial-0.49.1}/cellestial/util/utilities.py +15 -29
  24. {cellestial-0.48.4 → cellestial-0.49.1}/pyproject.toml +1 -1
  25. cellestial-0.48.4/cellestial/single/heatmap/_key_groups.py +0 -317
  26. cellestial-0.48.4/cellestial/single/heatmap/_rank_genes_groups.py +0 -187
  27. {cellestial-0.48.4 → cellestial-0.49.1}/LICENSE +0 -0
  28. {cellestial-0.48.4 → cellestial-0.49.1}/README.md +0 -0
  29. {cellestial-0.48.4 → cellestial-0.49.1}/cellestial/__init__.py +0 -0
  30. {cellestial-0.48.4 → cellestial-0.49.1}/cellestial/_version.py +0 -0
  31. {cellestial-0.48.4 → cellestial-0.49.1}/cellestial/datasets/__init__.py +0 -0
  32. {cellestial-0.48.4 → cellestial-0.49.1}/cellestial/frames/__init__.py +0 -0
  33. {cellestial-0.48.4 → cellestial-0.49.1}/cellestial/layers/__init__.py +0 -0
  34. {cellestial-0.48.4 → cellestial-0.49.1}/cellestial/layers/_deferred.py +0 -0
  35. {cellestial-0.48.4 → cellestial-0.49.1}/cellestial/layers/ondata_legend.py +0 -0
  36. {cellestial-0.48.4 → cellestial-0.49.1}/cellestial/layers/stream.py +0 -0
  37. {cellestial-0.48.4 → cellestial-0.49.1}/cellestial/layers/utilities.py +0 -0
  38. {cellestial-0.48.4 → cellestial-0.49.1}/cellestial/single/__init__.py +0 -0
  39. {cellestial-0.48.4 → cellestial-0.49.1}/cellestial/single/base/__init__.py +0 -0
  40. {cellestial-0.48.4 → cellestial-0.49.1}/cellestial/single/base/base.py +0 -0
  41. {cellestial-0.48.4 → cellestial-0.49.1}/cellestial/single/basic/__init__.py +0 -0
  42. {cellestial-0.48.4 → cellestial-0.49.1}/cellestial/single/basic/bar.py +0 -0
  43. {cellestial-0.48.4 → cellestial-0.49.1}/cellestial/single/basic/scatter.py +0 -0
  44. {cellestial-0.48.4 → cellestial-0.49.1}/cellestial/single/common/__init__.py +0 -0
  45. {cellestial-0.48.4 → cellestial-0.49.1}/cellestial/single/common/xyplot.py +0 -0
  46. {cellestial-0.48.4 → cellestial-0.49.1}/cellestial/single/common/xyplots.py +0 -0
  47. {cellestial-0.48.4 → cellestial-0.49.1}/cellestial/single/core/__init__.py +0 -0
  48. {cellestial-0.48.4 → cellestial-0.49.1}/cellestial/single/core/distributions.py +0 -0
  49. {cellestial-0.48.4 → cellestial-0.49.1}/cellestial/single/differential/__init__.py +0 -0
  50. {cellestial-0.48.4 → cellestial-0.49.1}/cellestial/single/heatmap/__init__.py +0 -0
  51. {cellestial-0.48.4 → cellestial-0.49.1}/cellestial/single/quick/__init__.py +0 -0
  52. {cellestial-0.48.4 → cellestial-0.49.1}/cellestial/single/quick/elbow.py +0 -0
  53. {cellestial-0.48.4 → cellestial-0.49.1}/cellestial/single/quick/highest.py +0 -0
  54. {cellestial-0.48.4 → cellestial-0.49.1}/cellestial/single/quick/ridge.py +0 -0
  55. {cellestial-0.48.4 → cellestial-0.49.1}/cellestial/spatial/__init__.py +0 -0
  56. {cellestial-0.48.4 → cellestial-0.49.1}/cellestial/spatial/spatials.py +0 -0
  57. {cellestial-0.48.4 → cellestial-0.49.1}/cellestial/themes/__init__.py +0 -0
  58. {cellestial-0.48.4 → cellestial-0.49.1}/cellestial/themes/_distributions.py +0 -0
  59. {cellestial-0.48.4 → cellestial-0.49.1}/cellestial/themes/_heatmaps.py +0 -0
  60. {cellestial-0.48.4 → cellestial-0.49.1}/cellestial/themes/_scatters.py +0 -0
  61. {cellestial-0.48.4 → cellestial-0.49.1}/cellestial/util/__init__.py +0 -0
  62. {cellestial-0.48.4 → cellestial-0.49.1}/cellestial/util/colors.py +0 -0
  63. {cellestial-0.48.4 → cellestial-0.49.1}/cellestial/util/errors.py +0 -0
  64. {cellestial-0.48.4 → cellestial-0.49.1}/cellestial/util/operations.py +0 -0
  65. {cellestial-0.48.4 → cellestial-0.49.1}/cellestial/util/save.py +0 -0
@@ -1,6 +1,6 @@
1
1
  Metadata-Version: 2.4
2
2
  Name: cellestial
3
- Version: 0.48.4
3
+ Version: 0.49.1
4
4
  Summary:
5
5
  License-Expression: Apache-2.0
6
6
  License-File: LICENSE
@@ -487,7 +487,7 @@ def from_url(
487
487
  print(f"No cache at {cache_file} !")
488
488
  print(f"Downloading from {url} ...")
489
489
 
490
- block_size = 1024
490
+ block_size = 1024 * 1024
491
491
  # Stream to a .part file and atomically rename on success, so an interrupted
492
492
  # download cannot poison the cache.
493
493
  partial_file = cache_file.with_name(cache_file.name + ".part")
@@ -2,9 +2,11 @@ from __future__ import annotations
2
2
 
3
3
  from typing import TYPE_CHECKING, Literal
4
4
 
5
+ import numpy as np
5
6
  import pandas as pd
6
7
  import polars as pl
7
8
  from anndata import AnnData
9
+ from scipy.sparse import issparse
8
10
  from spatialdata import SpatialData
9
11
 
10
12
  from cellestial.util.errors import UnsupportedDataTypeError, VariableNotFoundError
@@ -29,17 +31,23 @@ def anndata_variable_columns(
29
31
  columns = []
30
32
  if isinstance(keys, str):
31
33
  keys = [keys]
32
- for key in keys:
33
- if key in column_names: # check for repeats
34
- continue
35
- elif key in data.var_names:
36
- values = data.obs_vector(key)
37
- # add the variable column to the list of columns
38
- columns.append(pl.Series(key, values).cast(pl.Float32))
39
- column_names.append(key)
40
- else:
41
- msg = f"Key `{key}` not found in data."
42
- raise VariableNotFoundError(msg)
34
+ # remove keys that are already in column_names to avoid repeats
35
+ keys = [key for key in keys if key not in column_names]
36
+ # check that all keys are in data.var_names
37
+ available_names = set(data.var_names)
38
+ missing = [key for key in keys if key not in available_names]
39
+ if missing:
40
+ msg = f"Keys not found in data.var_names: {missing}"
41
+ raise VariableNotFoundError(msg)
42
+
43
+ # subset X
44
+ X = data[:, keys].X
45
+
46
+ if issparse(X):
47
+ X = X.toarray() # ty:ignore[unresolved-attribute]
48
+ else:
49
+ X = np.asarray(X)
50
+ columns = [pl.Series(key, X[:, i]) for i, key in enumerate(keys)]
43
51
 
44
52
  return columns
45
53
 
@@ -210,7 +218,7 @@ def anndata_variables_frame(
210
218
 
211
219
 
212
220
  def build_frame(
213
- data: AnnData,
221
+ data: AnnData | SpatialData,
214
222
  *,
215
223
  variable_keys: str | Sequence[str] | None = None,
216
224
  axis: Literal[0, 1] | None = None,
@@ -223,8 +231,8 @@ def build_frame(
223
231
 
224
232
  Parameters
225
233
  ----------
226
- data : AnnData
227
- The AnnData object containing the variables.
234
+ data : AnnData | SpatialData
235
+ The AnnData or SpatialData object containing the variables.
228
236
  variable_keys : str | Sequence[str] | None
229
237
  Variable keys to add to the DataFrame. If None, no additional keys are added.
230
238
  axis : {0,1} | None
@@ -282,6 +290,7 @@ def build_frame(
282
290
  if isinstance(data, SpatialData):
283
291
  tables = list(data.tables.keys())
284
292
  if len(tables) == 1:
293
+ # retrieve an anndata
285
294
  data = data.tables[tables[0]]
286
295
  elif len(tables) == 0:
287
296
  msg = "No annotation tables found in the SpatialData object."
@@ -6,7 +6,6 @@ from anndata import AnnData
6
6
  from polars import DataFrame
7
7
  from scipy.sparse import issparse
8
8
 
9
- from cellestial.frames.build import anndata_variable_columns
10
9
  from cellestial.util.errors import UnsupportedDataTypeError
11
10
 
12
11
 
@@ -16,28 +15,44 @@ def _highest_expressed_genes_frame(
16
15
  ) -> DataFrame:
17
16
  """Get the top n highest expressed genes by mean percentage across all cells."""
18
17
  if isinstance(data, AnnData):
18
+ # VALIDATE: inputs
19
19
  if n > data.n_vars:
20
20
  msg = f"Requested n={n} genes, but only {data.n_vars} genes available in data."
21
21
  raise ValueError(msg)
22
22
  X = data.X
23
- # normalize each cell to sum to 100 (percentage)
23
+ # matmul and fancy indexing below require an in-memory ndarray/sparse;
24
+ # fail fast on backed/dask/zappy rather than crashing deep in the math.
25
+ if not (isinstance(X, np.ndarray) or issparse(X)):
26
+ msg = (
27
+ f"Unsupported `X` type for this operation: {type(X).__name__}. "
28
+ "Load AnnData in non-backed mode or materialize `X` manually."
29
+ )
30
+ raise UnsupportedDataTypeError(msg)
31
+
32
+ # BUILD: per-cell normalizer so each cell's expression sums to 100
24
33
  if issparse(X):
25
- row_sums = np.array(X.sum(axis=1)).ravel()
26
- row_sums[row_sums == 0] = 1
27
- X_normalized = X.multiply(100 / row_sums[:, np.newaxis])
28
- mean_percent = np.array(X_normalized.mean(axis=0)).ravel()
34
+ row_sums = np.asarray(X.sum(axis=1)).ravel()
29
35
  else:
30
- row_sums = X.sum(axis=1, keepdims=True)
31
- row_sums[row_sums == 0] = 1
32
- X_normalized = X / row_sums * 100
33
- mean_percent = X_normalized.mean(axis=0)
36
+ row_sums = X.sum(axis=1)
37
+ row_sums[row_sums == 0] = 1 # guard cells with zero total to avoid div-by-zero
38
+ inv_rows = 100.0 / row_sums
34
39
 
40
+ # RANK: gene means via a single matvec instead of materializing the
41
+ # full normalized matrix. Identity: mean(X / row_sums * 100, axis=0)
42
+ # equals (inv_rows @ X) / n_cells, but uses O(n_vars) memory instead of O(n_obs * n_vars).
43
+ mean_percent = np.asarray(inv_rows @ X).ravel() / X.shape[0]
35
44
  top_idx = np.argsort(mean_percent)[::-1][:n]
36
45
  genes = data.var_names[top_idx].tolist()
37
- if issparse(X_normalized):
38
- X_normalized = X_normalized.tocsr()
39
- norm_data = AnnData(X=X_normalized, var=data.var, obs=data.obs)
40
- frame = pl.DataFrame(anndata_variable_columns(norm_data, keys=genes, column_names=[]))
46
+
47
+ # EXTRACT: normalize only the top n columns (small n_obs * n slice),
48
+ # skipping the full n_obs * n_vars materialization the naive path would do.
49
+ X_top = X[:, top_idx]
50
+ if issparse(X_top):
51
+ X_top = X_top.toarray()
52
+ X_top_normalized = X_top * inv_rows[:, np.newaxis]
53
+
54
+ # BUILD: polars frame keyed by gene name
55
+ frame = pl.DataFrame({gene: X_top_normalized[:, i] for i, gene in enumerate(genes)})
41
56
  else:
42
57
  msg = f"Unsupported data type: `{type(data)}`"
43
58
  raise UnsupportedDataTypeError(msg)
@@ -2,11 +2,14 @@ from __future__ import annotations
2
2
 
3
3
  from typing import TYPE_CHECKING, Literal
4
4
 
5
+ import polars as pl
5
6
  from lets_plot import (
7
+ aes,
6
8
  arrow,
7
9
  element_blank,
8
10
  element_text,
9
11
  geom_blank,
12
+ geom_path,
10
13
  geom_segment,
11
14
  theme,
12
15
  )
@@ -19,6 +22,111 @@ if TYPE_CHECKING:
19
22
  from lets_plot.plot.core import FeatureSpecArray, PlotSpec
20
23
  from polars import DataFrame
21
24
 
25
+ # fraction of each axis arm drawn as the arrowhead-bearing tip segment;
26
+ # lets_plot caps the arrowhead to ~half the segment length, so the tip must be
27
+ # long enough to render a full-size head
28
+ _TIP_FRACTION = 0.3
29
+
30
+
31
+ def _axis_arrow_layers(
32
+ frame: DataFrame,
33
+ *,
34
+ x: str,
35
+ y: str,
36
+ length: float,
37
+ size: float,
38
+ color: str,
39
+ angle: float,
40
+ arrow_kwargs: dict | None = None,
41
+ ) -> FeatureSpecArray:
42
+ """
43
+ Build the arrow-axis as an L-shaped path plus two arrowhead tip segments.
44
+
45
+ The semi-square L is one continuous `geom_path` that stops short of each
46
+ open end so the corner joins cleanly with no overlap. Each tip is completed
47
+ by a `geom_segment` carrying the arrowhead, meeting the path end-to-end.
48
+
49
+ Parameters
50
+ ----------
51
+ frame : `polars.DataFrame`
52
+ DataFrame holding the `x` and `y` columns.
53
+ x : str
54
+ Name of the x axis column.
55
+ y : str
56
+ Name of the y axis column.
57
+ length : float
58
+ Length of each axis arm as a fraction of its data range.
59
+ size : float
60
+ Line width of the path and tip segments.
61
+ color : str
62
+ Color of the path and tip segments.
63
+ angle : float
64
+ Angle of the arrowhead in degrees.
65
+ arrow_kwargs : dict | None
66
+ Additional parameters for the `arrow` layer.
67
+
68
+ Returns
69
+ -------
70
+ `FeatureSpecArray`
71
+ The path and tip-segment layers.
72
+ """
73
+ arrow_kwargs = arrow_kwargs or {}
74
+
75
+ x_min = frame[x].min()
76
+ x_max = frame[x].max()
77
+ y_min = frame[y].min()
78
+ y_max = frame[y].max()
79
+
80
+ # find total difference between the max and min for both axis
81
+ x_diff = x_max - x_min
82
+ y_diff = y_max - y_min
83
+
84
+ # find the ends of the arrows
85
+ xend = x_min + length * x_diff
86
+ yend = y_min + length * y_diff
87
+
88
+ # adjust bottom ends of arrows
89
+ adjust_rate = 0.025
90
+ x0 = x_min - x_diff * adjust_rate
91
+ y0 = y_min - y_diff * adjust_rate
92
+
93
+ # tip segments carry the arrowheads; the path stops short of each tip
94
+ tip_x = (xend - x0) * _TIP_FRACTION
95
+ tip_y = (yend - y0) * _TIP_FRACTION
96
+
97
+ # semi-square L (stopping short of both tips), drawn as one connected path
98
+ path_frame = pl.DataFrame(
99
+ {"x": [xend - tip_x, x0, x0], "y": [y0, y0, yend - tip_y]}
100
+ )
101
+
102
+ layers = geom_path(
103
+ data=path_frame,
104
+ mapping=aes(x="x", y="y"),
105
+ color=color,
106
+ size=size,
107
+ )
108
+ # X axis tip
109
+ layers += geom_segment(
110
+ x=xend - tip_x,
111
+ y=y0,
112
+ xend=xend,
113
+ yend=y0,
114
+ color=color,
115
+ size=size,
116
+ arrow=arrow(angle, **arrow_kwargs),
117
+ )
118
+ # Y axis tip
119
+ layers += geom_segment(
120
+ x=x0,
121
+ y=yend - tip_y,
122
+ xend=x0,
123
+ yend=yend,
124
+ color=color,
125
+ size=size,
126
+ arrow=arrow(angle, **arrow_kwargs),
127
+ )
128
+ return layers
129
+
22
130
 
23
131
  def _modify_axis(
24
132
  frame: DataFrame,
@@ -45,13 +153,13 @@ def _modify_axis(
45
153
  axis_type : {'axis', 'arrow'} | None
46
154
  Whether to use regular axis or arrows as the axis.
47
155
  arrow_size : float
48
- Size of the arrow.
156
+ Width of the axis lines.
49
157
  arrow_color : str
50
158
  Color of the arrow.
51
159
  arrow_angle : float
52
160
  Angle of the arrow head in degrees.
53
161
  arrow_length : float
54
- Length of the arrow head (px).
162
+ Span of each axis line as a fraction of its data range (0.25 covers 25%).
55
163
 
56
164
  Returns
57
165
  -------
@@ -86,44 +194,14 @@ def _modify_axis(
86
194
  axis_title_x=element_text(hjust=arrow_length / 2.5), # better than 2
87
195
  axis_title_y=element_text(hjust=arrow_length / 2.5),
88
196
  )
89
- x_max = frame[x].max()
90
- x_min = frame[x].min()
91
- y_max = frame[y].max()
92
- y_min = frame[y].min()
93
-
94
- # find total difference between the max and min for both axis
95
- x_diff = x_max - x_min # ty:ignore[unsupported-operator]
96
- y_diff = y_max - y_min # ty:ignore[unsupported-operator]
97
-
98
- # find the ends of the arrows
99
- # ensure the arrow length is the same for both axis
100
- xend = x_min + arrow_length * x_diff
101
- yend = y_min + arrow_length * y_diff
102
-
103
- # adjust bottom ends of arrows
104
- adjust_rate = 0.025
105
- x0 = x_min - x_diff * adjust_rate
106
- y0 = y_min - y_diff * adjust_rate
107
-
108
- # X axis
109
- new_layer += geom_segment(
110
- x=x0,
111
- y=y0,
112
- xend=xend,
113
- yend=y0,
114
- color=arrow_color,
197
+ new_layer += _axis_arrow_layers(
198
+ frame,
199
+ x=x,
200
+ y=y,
201
+ length=arrow_length,
115
202
  size=arrow_size,
116
- arrow=arrow(arrow_angle),
117
- )
118
- # Y axis
119
- new_layer += geom_segment(
120
- x=x0,
121
- y=y0,
122
- xend=x0,
123
- yend=yend,
124
203
  color=arrow_color,
125
- size=arrow_size,
126
- arrow=arrow(arrow_angle),
204
+ angle=arrow_angle,
127
205
  )
128
206
  else:
129
207
  msg = f"expected 'axis' or 'arrow' for 'axis_type' argument, but received {axis_type}"
@@ -151,13 +229,13 @@ def arrow_axis(
151
229
  regardless of which plot the resulting layer is added to. When `None`,
152
230
  the layer is deferred and introspects the plot it is added to via `+`.
153
231
  size : float
154
- Size of the arrow.
232
+ Width of the axis lines.
155
233
  color : str
156
234
  Color of the arrow.
157
235
  angle : float
158
236
  Angle of the arrow head in degrees.
159
237
  length : float
160
- Length of the arrow head (px).
238
+ Span of each axis line as a fraction of its data range (0.25 covers 25%).
161
239
  **arrow_kwargs
162
240
  Additional parameters for the `arrow` layer.
163
241
  For more information on arrow parameters, see:
@@ -242,43 +320,15 @@ def arrow_axis(
242
320
  axis_title_x=element_text(hjust=length / 2.5), # better than 2
243
321
  axis_title_y=element_text(hjust=length / 2.5),
244
322
  )
245
- x_max = frame[x].max()
246
- x_min = frame[x].min()
247
- y_max = frame[y].max()
248
- y_min = frame[y].min()
249
-
250
- # find total difference between the max and min for both axis
251
- x_diff = x_max - x_min
252
- y_diff = y_max - y_min
253
-
254
- # find the ends of the arrows
255
- xend = x_min + length * x_diff
256
- yend = y_min + length * y_diff
257
-
258
- # adjust bottom ends of arrows
259
- adjust_rate = 0.025
260
- x0 = x_min - x_diff * adjust_rate
261
- y0 = y_min - y_diff * adjust_rate
262
-
263
- # X axis
264
- new_layer += geom_segment(
265
- x=x0,
266
- y=y0,
267
- xend=xend,
268
- yend=y0,
269
- color=color,
323
+ new_layer += _axis_arrow_layers(
324
+ frame,
325
+ x=x,
326
+ y=y,
327
+ length=length,
270
328
  size=size,
271
- arrow=arrow(angle, **arrow_kwargs),
272
- )
273
- # Y axis
274
- new_layer += geom_segment(
275
- x=x0,
276
- y=y0,
277
- xend=x0,
278
- yend=yend,
279
329
  color=color,
280
- size=size,
281
- arrow=arrow(angle, **arrow_kwargs),
330
+ angle=angle,
331
+ arrow_kwargs=arrow_kwargs,
282
332
  )
283
333
  return new_layer
284
334
 
@@ -6,6 +6,7 @@ from typing import TYPE_CHECKING, Literal
6
6
  import numpy as np
7
7
  import polars as pl
8
8
  from lets_plot import aes, geom_bracket
9
+ from scipy.stats import mannwhitneyu, ttest_ind
9
10
 
10
11
  from cellestial.layers._deferred import DeferredLayer
11
12
  from cellestial.util import get_mapping, retrieve
@@ -134,8 +135,6 @@ def _compute_bracket_frame(
134
135
  y_padding: float,
135
136
  ) -> DataFrame:
136
137
  """Build a `geom_bracket` DataFrame from pairwise significance tests."""
137
- from scipy.stats import mannwhitneyu, ttest_ind
138
-
139
138
  # determine which pairs to compare
140
139
  groups = frame[x].drop_nulls().unique(maintain_order=True).to_list()
141
140
  if comparisons is None:
@@ -238,10 +237,10 @@ def _compute_bracket_frame(
238
237
  brackets = brackets.with_columns(label=pl.Series(labels))
239
238
 
240
239
  # compute y positions so brackets stack above the data without overlapping
241
- values = frame[y].to_numpy()
242
- data_min = float(values.min())
243
- data_max = float(values.max())
244
- data_range = data_max - data_min
240
+
241
+ data_min = frame[y].min()
242
+ data_max = frame[y].max()
243
+ data_range = data_max - data_min # ty:ignore[unsupported-operator]
245
244
  _y_position = data_max + data_range * y_padding if y_position is None else y_position
246
245
  _y_step = data_range * y_padding if y_step is None else y_step
247
246
  positions = [_y_position + _y_step * index for index in range(len(brackets))]
@@ -5,6 +5,7 @@ from typing import TYPE_CHECKING, Literal
5
5
  import numpy as np
6
6
  import polars as pl
7
7
  from lets_plot import aes, geom_path
8
+ from scipy.stats import gaussian_kde
8
9
 
9
10
  from cellestial.layers._deferred import DeferredLayer
10
11
  from cellestial.util import get_mapping
@@ -28,7 +29,6 @@ def _get_density_boundaries(
28
29
  grid_size: int = 200,
29
30
  ) -> DataFrame:
30
31
  """Creates and Returns a DataFrame to encircle the cluster via `geom_path`."""
31
- from scipy.stats import gaussian_kde
32
32
  from skimage import measure
33
33
 
34
34
  # 1. Normalize input to a Sequence of Sequences: List[List[str]]
@@ -116,9 +116,9 @@ def dimensional(
116
116
  axis_type : {'axis', 'arrow'} | None
117
117
  Whether to use regular axis or arrows as the axis.
118
118
  arrow_length : float, default=0.25
119
- Length of the arrow head (px).
119
+ Span of each axis line as a fraction of its data range (0.25 covers 25%).
120
120
  arrow_size : float, default=1
121
- Size of the arrow.
121
+ Width of the axis lines.
122
122
  arrow_color : str, default='#3f3f3f'
123
123
  Color of the arrows.
124
124
  arrow_angle : float, default=10