1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60 61 62 63 64 65 66 67 68 69 70 71 72 73 74 75 76 77 78 79 80 81 82 83 84 85 86 87 88 89 90 91 92 93 94 95 96 97 98 99 100 101 102 103 104 105 106 107 108 109 110 111 112 113 114 115 116 117 118 119 120 121 122 123 124 125 126 127 128 129 130 131 132 133 134 135 136 137 138 139 140 141 142 143 144 145 146 147 148 149 150 151 152 153 154 155 156 157 158 159 160 161 162 163 164 165 166 167 168 169 170 171 172 173 174 175 176 177 178 179 180 181 182 183 184 185 186 187 188 189 190 191 192 193 194 195 196 197 198 199 200 201 202 203 204 205 206 207 208 209 210 211 212 213 214 215 216 217 218 219 220 221 222 223 224 225 226 227 228 229 230 231 232 233 234 235 236 237 238 239 240 241 242 243 244 245 246 247 248 249 250 251 252 253 254 255 256 257 258 259 260 261 262 263 264 265 266 267 268 269 270 271 272 273 274 275 276 277 278 279 280 281 282 283 284 285 286 287 288 289 290 291 292 293 294 295 296 297 298 299 300 301 302 303 304 305 306 307 308 309 310 311 312 313 314 315 316 317 318 319 320 321 322 323 324 325 326 327 328 329 330 331 332 333 334 335 336 337 338 339 340 341 342 343 344 345 346 347 348 349
|
from abc import ABC, abstractmethod
from collections import abc as cabc
from typing import Union, Optional, Type, ClassVar, TypeVar # Special types
from typing import Iterator, Mapping, Sequence # ABCs
from typing import Tuple, List, Dict # Generic base types
import numpy as np
import pandas as pd
from scipy.sparse import spmatrix
from ..utils import deprecated, ensure_df_homogeneous
from . import raw, anndata
from .views import as_view
from .access import ElementRef
from .index import _subset
OneDIdx = Union[Sequence[int], Sequence[bool], slice]
TwoDIdx = Tuple[OneDIdx, OneDIdx]
I = TypeVar("I", OneDIdx, TwoDIdx, covariant=True)
# TODO: pd.DataFrame only allowed in AxisArrays?
V = Union[pd.DataFrame, spmatrix, np.ndarray]
class AlignedMapping(cabc.MutableMapping, ABC):
"""\
An abstract base class for Mappings containing array-like values aligned
to either one or both AnnData axes.
"""
_allow_df: ClassVar[bool]
"""If this mapping supports heterogeneous DataFrames"""
_view_class: ClassVar[Type["AlignedViewMixin"]]
"""The view class for this aligned mapping."""
_actual_class: ClassVar[Type["AlignedActualMixin"]]
"""The actual class (which has it’s own data) for this aligned mapping."""
def __repr__(self):
return f"{type(self).__name__} with keys: {', '.join(self.keys())}"
def _ipython_key_completions_(self) -> List[str]:
return list(self.keys())
def _validate_value(self, val: V, key: str) -> V:
"""Raises an error if value is invalid"""
for i, axis in enumerate(self.axes):
if self.parent.shape[axis] != val.shape[i]:
right_shape = tuple(self.parent.shape[a] for a in self.axes)
raise ValueError(
f"Value passed for key {key!r} is of incorrect shape. "
f"Values of {self.attrname} must match dimensions "
f"{self.axes} of parent. Value had shape {val.shape} while "
f"it should have had {right_shape}."
)
if not self._allow_df and isinstance(val, pd.DataFrame):
name = self.attrname.title().rstrip("s")
val = ensure_df_homogeneous(val, f"{name} {key!r}")
return val
@property
@abstractmethod
def attrname(self) -> str:
"""What attr for the AnnData is this?"""
pass
@property
@abstractmethod
def axes(self) -> Tuple[int, ...]:
"""Which axes of the parent is this aligned to?"""
pass
@property
@abstractmethod
def is_view(self) -> bool:
pass
@property
def parent(self) -> Union["anndata.AnnData", "raw.Raw"]:
return self._parent
def copy(self):
d = self._actual_class(self.parent, self._axis)
for k, v in self.items():
d[k] = v.copy()
return d
def _view(self, parent: "anndata.AnnData", subset_idx: I):
"""Returns a subset copy-on-write view of the object."""
return self._view_class(self, parent, subset_idx)
@deprecated("dict(obj)")
def as_dict(self) -> dict:
return dict(self)
class AlignedViewMixin:
parent: "anndata.AnnData"
"""Reference to parent AnnData view"""
attrname: str
"""What attribute in the parent is this?"""
parent_mapping: Mapping[str, V]
"""The object this is a view of."""
is_view = True
def __getitem__(self, key: str) -> V:
return as_view(
_subset(self.parent_mapping[key], self.subset_idx),
ElementRef(self.parent, self.attrname, (key,)),
)
def __setitem__(self, key: str, value: V):
value = self._validate_value(value, key) # Validate before mutating
adata = self.parent.copy()
new_mapping = getattr(adata, self.attrname)
new_mapping[key] = value
self.parent._init_as_actual(adata)
def __delitem__(self, key: str):
self[key] # Make sure it exists before bothering with a copy
adata = self.parent.copy()
new_mapping = getattr(adata, self.attrname)
del new_mapping[key]
self.parent._init_as_actual(adata)
def __contains__(self, key: str) -> bool:
return key in self.parent_mapping
def __iter__(self) -> Iterator[str]:
return iter(self.parent_mapping)
def __len__(self) -> int:
return len(self.parent_mapping)
class AlignedActualMixin:
_data: Dict[str, V]
"""Underlying mapping to the data"""
is_view = False
def __getitem__(self, key: str) -> V:
return self._data[key]
def __setitem__(self, key: str, value: V):
value = self._validate_value(value, key)
self._data[key] = value
def __contains__(self, key: str) -> bool:
return key in self._data
def __delitem__(self, key: str):
del self._data[key]
def __iter__(self) -> Iterator[str]:
return iter(self._data)
def __len__(self) -> int:
return len(self._data)
class AxisArraysBase(AlignedMapping):
"""\
Mapping of key→array-like,
where array-like is aligned to an axis of parent AnnData.
"""
_allow_df = True
_dimnames = ("obs", "var")
@property
def attrname(self) -> str:
return f"{self.dim}m"
@property
def axes(self) -> Tuple[int]:
"""Axes of the parent this is aligned to"""
return (self._axis,)
@property
def dim(self) -> str:
"""Name of the dimension this aligned to."""
return self._dimnames[self._axis]
def flipped(self) -> "AxisArraysBase":
"""Transpose."""
new = self.copy()
new.dimension = abs(self._axis - 1)
return new
def to_df(self) -> pd.DataFrame:
"""Convert to pandas dataframe."""
df = pd.DataFrame(index=self.dim_names)
for key in self.keys():
value = self[key]
for icolumn, column in enumerate(value.T):
df[f"{key}{icolumn + 1}"] = column
return df
def _validate_value(self, val: V, key: str) -> V:
if (
hasattr(val, "index")
and isinstance(val.index, cabc.Collection)
and not (val.index == self.dim_names).all()
):
# Could probably also re-order index if it’s contained
raise ValueError(
f"value.index does not match parent’s axis {self.axes[0]} names"
)
return super()._validate_value(val, key)
class AxisArrays(AlignedActualMixin, AxisArraysBase):
def __init__(
self,
parent: Union["anndata.AnnData", "raw.Raw"],
axis: int,
vals: Union[Mapping, AxisArraysBase, None] = None,
):
self._parent = parent
if axis not in (0, 1):
raise ValueError()
self._axis = axis
self.dim_names = (parent.obs_names, parent.var_names)[self._axis]
self._data = dict()
if vals is not None:
self.update(vals)
class AxisArraysView(AlignedViewMixin, AxisArraysBase):
def __init__(
self,
parent_mapping: AxisArraysBase,
parent_view: "anndata.AnnData",
subset_idx: OneDIdx,
):
self.parent_mapping = parent_mapping
self._parent = parent_view
self.subset_idx = subset_idx
self._axis = parent_mapping._axis
self.dim_names = parent_mapping.dim_names[subset_idx]
AxisArraysBase._view_class = AxisArraysView
AxisArraysBase._actual_class = AxisArrays
class LayersBase(AlignedMapping):
"""\
Mapping of key: array-like, where array-like is aligned to both axes of the
parent anndata.
"""
_allow_df = False
attrname = "layers"
axes = (0, 1)
# TODO: I thought I had a more elegant solution to overiding this...
def copy(self) -> "Layers":
d = self._actual_class(self.parent)
for k, v in self.items():
d[k] = v.copy()
return d
class Layers(AlignedActualMixin, LayersBase):
def __init__(self, parent: "anndata.AnnData", vals: Optional[Mapping] = None):
self._parent = parent
self._data = dict()
if vals is not None:
self.update(vals)
class LayersView(AlignedViewMixin, LayersBase):
def __init__(
self,
parent_mapping: LayersBase,
parent_view: "anndata.AnnData",
subset_idx: TwoDIdx,
):
self.parent_mapping = parent_mapping
self._parent = parent_view
self.subset_idx = subset_idx
LayersBase._view_class = LayersView
LayersBase._actual_class = Layers
class PairwiseArraysBase(AlignedMapping):
"""\
Mapping of key: array-like, where both axes of array-like are aligned to
one axis of the parent anndata.
"""
_allow_df = False
_dimnames = ("obs", "var")
@property
def attrname(self) -> str:
return f"{self.dim}p"
@property
def axes(self) -> Tuple[int, int]:
"""Axes of the parent this is aligned to"""
return self._axis, self._axis
@property
def dim(self) -> str:
"""Name of the dimension this aligned to."""
return self._dimnames[self._axis]
class PairwiseArrays(AlignedActualMixin, PairwiseArraysBase):
def __init__(
self,
parent: "anndata.AnnData",
axis: int,
vals: Optional[Mapping] = None,
):
self._parent = parent
if axis not in (0, 1):
raise ValueError()
self._axis = axis
self._data = dict()
if vals is not None:
self.update(vals)
class PairwiseArraysView(AlignedViewMixin, PairwiseArraysBase):
def __init__(
self,
parent_mapping: PairwiseArraysBase,
parent_view: "anndata.AnnData",
subset_idx: OneDIdx,
):
self.parent_mapping = parent_mapping
self._parent = parent_view
self.subset_idx = (subset_idx, subset_idx)
self._axis = parent_mapping._axis
PairwiseArraysBase._view_class = PairwiseArraysView
PairwiseArraysBase._actual_class = PairwiseArrays
|