iam-git / WellMet (public) (License: MIT) (since 2021-08-31) (hash sha1)
WellMet is pure Python framework for spatial structural reliability analysis. Or, more specifically, for "failure probability estimation and detection of failure surfaces by adaptive sequential decomposition of the design domain".
Clone URLs: https://rocketgit.com/user/iam-git/WellMet ssh://rocketgit@ssh.rocketgit.com/user/iam-git/WellMet git://git.rocketgit.com/user/iam-git/WellMet
release trunk
List of commits:
Subject Hash Author Date (UTC)
voronoi: implement faster TwoPoints class. c0af78f64bc78a6b8cf06e3ea244339549318b66 I am 2022-10-24 13:03:26
voronoi: implement TwoHorses class. Třída je, bohužel, bezcenná :( 4f6fe186f3f344fa7ee9e518798499099615b466 I am 2022-10-23 10:00:43
voronoi: WIP 23c63ede5bc2ec51d25aed850dc59299db438e34 I am 2022-10-21 12:27:01
IS_stat.sample_alike: use any possible fallback options to sample 80fb6d086e9f17350a1725a319e40bf29591869e I am 2022-10-20 08:56:42
voronoi: WIP ba912d406de12142f2e15a73090118184a0166a4 I am 2022-10-20 02:43:12
IS_stat: add sample_alike() function 2e5394d264b7aaaf5f2486c736ed8cc37afbd030 I am 2022-10-19 13:00:17
voronoi: WIP c3bfa60b6db2042df11528fe95a51d7cc88d3c6b I am 2022-10-19 11:18:13
voronoi: WIP baff3c5ccaf857f60194917fe5d6e33c40b01265 I am 2022-10-19 09:28:33
voronoi.ConvexSpline: little fix of np.sign() for better sleep 968f649e54d97abd596cd60749872fa3ef788c06 I am 2022-10-19 06:28:46
implement even more optimized Convex.. now ConvexSpline class c08511aecb32b880e73c9f10d07feff4dc588e8b I am 2022-10-19 04:11:13
voronoi: WIP cbb61d4544367a67c9a5c3fe6d64a9dc284eb364 I am 2022-10-18 05:05:44
voronoi: WIP e9c247fef30cfb678bad9b955d6f6a94a0ff61e7 I am 2022-10-18 01:02:30
candynodex: hotfix to support numpy masking indexing f359123b72a3d998d5eed4b9e611a5e402a18c8d I am 2022-10-17 09:02:33
add new lightweight candynodes module to replace old heavy ugly candybox ca97509f47c100df90e1e06d82ed7c759627bfd0 I am 2022-10-17 06:31:33
qt_gui: move CandyBox creating to the individual DiceBox widgets 29253625129219b5e550f82cae3da0e53ef5bd47 I am 2022-10-16 03:17:55
voronoi: WIP 4855190869e06df740cc05c64e8b27b2c3c5a88d I am 2022-10-15 22:46:51
voronoi: WIP 0298b3a0563587ea4fc24fa99b51d2e01d517203 I am 2022-10-15 20:00:25
voronoi: WIP 4b80757fa26e3d40dceac888480ec3d814f6abc9 I am 2022-10-15 17:30:11
voronoi: WIP 7cadd1d01d7f4bc3f51e785fe6ebf8cae6ff839a I am 2022-10-15 03:12:54
voronoi: clean up and comment out ContactSolver 543f8482a925ff6cf8485b9be616b5bd3c714d13 I am 2022-10-14 04:24:40
Commit c0af78f64bc78a6b8cf06e3ea244339549318b66 - voronoi: implement faster TwoPoints class.
Author: I am
Author date (UTC): 2022-10-24 13:03
Committer name: I am
Committer date (UTC): 2022-10-24 13:03
Parent(s): 4f6fe186f3f344fa7ee9e518798499099615b466
Signer:
Signing key:
Signing status: N
Tree: ac6f95f23a1613fff34711133fdee58daee54ede
File Lines added Lines deleted
wellmet/voronoi.py 194 0
File wellmet/voronoi.py changed (mode: 100644) (index 470aebb..4554c99)
... ... import scipy.stats as stats
8 8 from scipy.spatial import KDTree from scipy.spatial import KDTree
9 9 from scipy.spatial import Delaunay from scipy.spatial import Delaunay
10 10 from scipy import spatial from scipy import spatial
11 from scipy.spatial.distance import cdist
11 12 from scipy import interpolate from scipy import interpolate
12 13
13 14 import collections # for defaultdict, namedtuple import collections # for defaultdict, namedtuple
 
... ... class TwoHorses:
513 514
514 515
515 516
517 class TwoPoints:
518 def __init__(self, points, couple_indices):
519 i, j = couple_indices
520
521 self.point_i = point_i = points[i]
522 self.point_j = point_j = points[j]
523 #č vektor zadáme přímkou mezí vzorky
524 #č od "i" k "j", normalizovat nebudeme
525 self.normal_vector = point_j - point_i
526
527 #č jako indice zkusme použit bod na usečce uprostřed mezí vzorky
528 self.half_point = np.mean(points[[i, j]], axis=0)
529
530 self.delimit = np.inner(self.half_point, self.normal_vector)
531
532 #č pokud skalarní součin je menší jak delimit
533 #č (protože směr normály od "i" k "j"),
534 #č tak patří hledané do poloprostoru "i"
535 mask = (points @ self.normal_vector) < self.delimit
536
537
538 self.tree_i = KDTree(points[mask])
539 self.tree_j = KDTree(points[~mask])
540 #č neznám convinient cestu pro získaní
541 #č tamtamtoho indexu na maskovaném poli
542 #č np.ma mně příjde zbytečným
543 # ha! monkey patching
544 self.tree_i.idx = len(mask[:i][mask[:i]])
545 self.tree_j.idx = len(mask[:j][~mask[:j]])
546
547 #č bacha na indexy. Musíme vyhodit i-tý a j-tý vzorky
548 mask[i] = False
549 self.points_i = points[mask]
550 #č teď ale jedeme se stejnou, ale invertovanou maskou
551 mask[i] = True
552 mask[j] = True
553 self.points_j = points[~mask]
554
555
556 def query(self, nodes, order='F'):
557 #č pokud skalarní součin je menší jak delimit
558 #č (protože směr normály od "i" k "j"),
559 #č tak patří hledané do poloprostoru "i"
560 halfmask = (nodes @ self.normal_vector) < self.delimit
561
562 nodes_i = nodes[halfmask]
563 nodes_j = nodes[~halfmask]
564
565
566 status_i = status_j = False
567 if nodes_i.size:
568 status_i, mask_i, dd_i = self._run(self.point_j, self.points_i,
569 self.tree_j, nodes_i, order)
570
571 if nodes_j.size:
572 status_j, mask_j, dd_j = self._run(self.point_i, self.points_j,
573 self.tree_i, nodes_j, order)
574
575 if not (status_i or status_j):
576 return False, None, None
577
578 #č globální maska
579 gmask = np.empty(len(nodes), dtype=bool)
580
581 if status_i and not status_j:
582 gmask[halfmask] = mask_i
583 gmask[~halfmask] = False
584 return True, gmask, TData(X=nodes[gmask], imask=mask_i[mask_i], dd=dd_i)
585
586 if status_j and not status_i:
587 gmask[halfmask] = False
588 gmask[~halfmask] = mask_j
589 return True, gmask, TData(X=nodes[gmask], imask=~mask_j[mask_j], dd=dd_j)
590
591
592 gmask[halfmask] = mask_i
593 gmask[~halfmask] = mask_j
594
595 dd_mask = halfmask[gmask]
596 dd = np.empty(len(dd_i) + len(dd_j))
597 dd[dd_mask] = dd_i
598 dd[~dd_mask] = dd_j
599
600 return True, gmask, TData(X=nodes[gmask], imask=dd_mask, dd=dd)
601
602
603 @staticmethod
604 def _run(center_2, points_1, tree_2, nodes_1, order):
605 """
606 č Úkolem první fáze je vyhodit co nejvíc bodíků
607 č Bodíky bereme z i-tého poloprostoru a kontrolní vzdálenosti -
608 č vůči vzorku "j" z protílehlé strany.
609 č Předpokládáme, že na konci první fázi získame už
610 č dobře vyfiltrované tečičky, které přece pro formu
611 č musíme zkontrolovat i vůči vzorkům na opačné straně.
612 č Předpokladáme, že zůstaly jíž vícemené dobré bodíky
613 č a že smyčka druhé fáze už by musela běžet přes skoro všechno.
614 č Proto se nebudeme snažit dohnat Cečkovej kód v Pythonu
615 č a poslední část pokorně svěříme KDTree -
616 č nejrychlejší implementaci the nearest search na dívokém západě.
617 """
618 #č cdist vysloveně chce 2D pola, viděl jsem to i v kódě.
619 #č Vrací sloupcovou matici. Co kdyby s F pořadí by to jelo rychlejc?
620 #d1 = np.empty((len(nodes), 1), dtype=float, order=order)
621 #cdist(nodes, [center_1], 'sqeuclidean', out=d1)
622 d2 = np.empty((len(nodes_1), 1), dtype=float, order=order)
623 cdist(nodes_1, [center_2], 'sqeuclidean', out=d2)
624
625 mask = np.isfinite(d2)
626
627 # #č udělám to takhle i když neumím si představit,
628 # #č jak se můžou objevit NaNs a nekonečna ve d2, když nejsou v d1
629 # mask = np.isfinite(d1) & np.isfinite(d2)
630
631 n = len(mask[mask])
632 d_i = np.empty((n, 1), dtype=float, order=order)
633
634 #č pro nás spíše budou kritické poslední přídané vzorky
635 for point in reversed(points_1):
636 cdist(nodes_1[mask[:,0]], np.atleast_2d(point), 'sqeuclidean', out=d_i)
637
638 mask[mask[:,0]] &= d2[mask[:,0]] <= d_i
639
640 n = len(mask[mask])
641 if n == 0:
642 return False, None, None #mask[:,0], d_2[mask_2]
643 d_i.resize((n, 1), refcheck=False)
644
645 #č ty špinavé 2D matice zhora jíž nepotřebujem
646 #č převedeme (co nejlevnějc) masku na 1D
647 mask = mask[:,0]
648 d_2, j = tree_2.query(nodes_1[mask], k=1, p=2, workers=-1)
649 mask_2 = (j == tree_2.idx)
650
651 if np.any(mask_2):
652 #č hotovo. Vzdálenosti jíž nemusíme kontrolovat.
653 #č ve smyčce d2 jsme jíž prohnali.
654 mask[mask] = mask_2
655 return True, mask, d_2[mask_2]
656 else:
657 return False, None, None
658
659
660
661 # @staticmethod
662 # def _first_query(center, points, nodes, order):
663 #
664 # d1 = np.empty((len(nodes), 1), dtype=float, order=order)
665 # d2 = np.full((len(nodes), 1), np.inf, dtype=float, order=order)
666 # cdist(nodes, [center], 'sqeuclidean', out=d1)
667 # mask = np.isfinite(d1)
668 #
669 # n = len(mask[mask])
670 # #d2 = np.full((n, 1), np.inf, dtype=float, order=order)
671 # d3 = np.empty((n, 1), dtype=float, order=order)
672 #
673 # for point in reversed(points):
674 # cdist(nodes[mask[:,0]], np.atleast_2d(point), 'sqeuclidean', out=d3)
675 #
676 # np.fmin(d2[mask[:,0]], d3, out=d2[mask[:,0]])
677 # mask[mask[:,0]] &= d1[mask[:,0]] <= d3
678 #
679 # n = len(mask[mask])
680 # if n == 0:
681 # break
682 # d3.resize((n, 1), refcheck=False)
683 # return mask, d1, d2
684 #
685 # @staticmethod
686 # def _second_query(center, points, nodes, order):
687 #
688 # d1 = np.empty((len(nodes), 1), dtype=float, order=order)
689 # cdist(nodes, [center], 'sqeuclidean', out=d1)
690 # mask = np.isfinite(d1)
691 # n = len(mask[mask])
692 # d3 = np.empty((n, 1), dtype=float, order=order)
693 #
694 # for point in reversed(points):
695 # cdist(nodes[mask[:,0]], np.atleast_2d(point), 'sqeuclidean', out=d3)
696 #
697 # mask[mask[:,0]] &= d1[mask[:,0]] <= d3
698 #
699 # n = len(mask[mask])
700 # if n == 0:
701 # break
702 # d3.resize((n, 1), refcheck=False)
703 # return mask, d1
704
705
706
707
708
709
516 710
517 711 #оӵ кык точкаен Вороной #оӵ кык точкаен Вороной
518 712 # KickTouchCayenneVoronoi # KickTouchCayenneVoronoi
Hints:
Before first commit, do not forget to setup your git environment:
git config --global user.name "your_name_here"
git config --global user.email "your@email_here"

Clone this repository using HTTP(S):
git clone https://rocketgit.com/user/iam-git/WellMet

Clone this repository using ssh (do not forget to upload a key first):
git clone ssh://rocketgit@ssh.rocketgit.com/user/iam-git/WellMet

Clone this repository using git:
git clone git://git.rocketgit.com/user/iam-git/WellMet

You are allowed to anonymously push to this repository.
This means that your pushed commits will automatically be transformed into a merge request:
... clone the repository ...
... make some changes and some commits ...
git push origin main