Create a two-dimensional array with the flattened input as a diagonal. Parameters ---------- v : array_like Input data, which is flattened and set as the `k`-th diagonal of the output. k : int, optional Diagonal to set; 0, the default, corresponds to the
(v, k=0)
| 332 | |
| 333 | @array_function_dispatch(_diag_dispatcher) |
| 334 | def diagflat(v, k=0): |
| 335 | """ |
| 336 | Create a two-dimensional array with the flattened input as a diagonal. |
| 337 | |
| 338 | Parameters |
| 339 | ---------- |
| 340 | v : array_like |
| 341 | Input data, which is flattened and set as the `k`-th |
| 342 | diagonal of the output. |
| 343 | k : int, optional |
| 344 | Diagonal to set; 0, the default, corresponds to the "main" diagonal, |
| 345 | a positive (negative) `k` giving the number of the diagonal above |
| 346 | (below) the main. |
| 347 | |
| 348 | Returns |
| 349 | ------- |
| 350 | out : ndarray |
| 351 | The 2-D output array. |
| 352 | |
| 353 | See Also |
| 354 | -------- |
| 355 | diag : MATLAB work-alike for 1-D and 2-D arrays. |
| 356 | diagonal : Return specified diagonals. |
| 357 | trace : Sum along diagonals. |
| 358 | |
| 359 | Examples |
| 360 | -------- |
| 361 | >>> import numpy as np |
| 362 | >>> np.diagflat([[1,2], [3,4]]) |
| 363 | array([[1, 0, 0, 0], |
| 364 | [0, 2, 0, 0], |
| 365 | [0, 0, 3, 0], |
| 366 | [0, 0, 0, 4]]) |
| 367 | |
| 368 | >>> np.diagflat([1,2], 1) |
| 369 | array([[0, 1, 0], |
| 370 | [0, 0, 2], |
| 371 | [0, 0, 0]]) |
| 372 | |
| 373 | """ |
| 374 | conv = _array_converter(v) |
| 375 | v, = conv.as_arrays(subok=False) |
| 376 | v = v.ravel() |
| 377 | s = len(v) |
| 378 | n = s + abs(k) |
| 379 | res = zeros((n, n), v.dtype) |
| 380 | if (k >= 0): |
| 381 | i = arange(0, n - k, dtype=intp) |
| 382 | fi = i + k + i * n |
| 383 | else: |
| 384 | i = arange(0, n + k, dtype=intp) |
| 385 | fi = i + (i - k) * n |
| 386 | res.flat[fi] = v |
| 387 | |
| 388 | return conv.wrap(res) |
| 389 | |
| 390 | @finalize_array_function_like |
| 391 | @set_module('numpy') |
searching dependent graphs…