1 | | ~~Replying to [comment:40 jdemeyer]:~~ |

2 | | ~~> Replying to [comment:39 nthiery]:~~ |

3 | | ~~> > Please do not change the enumeration order, at least as default~~ |

4 | | ~~> I disagree with this: the default should be "do not sort, return stuff in the fastest possible way". Sorting an iterator is very expensive and should only be done if really needed.~~ |

5 | | ~~> ~~ |

6 | | ~~> > quite some code depends on it~~ |

7 | | ~~> Is that really true? The only doctest failures that I saw where "obvious" failures where some list order changed, I didn't see anything subtle.~~ |

8 | | |

9 | | ~~What about~~ |

10 | | ~~{{{~~ |

11 | | ~~sage: spc = SymmetricGroupRepresentation([2,2,1], 'orthogonal')~~ |

12 | | ~~sage: spc([1,3,2,4,5])~~ |

13 | | ~~[ -1/2 1/2*sqrt(3) 0 0 0]~~ |

14 | | ~~[1/2*sqrt(3) 1/2 0 0 0]~~ |

15 | | ~~[ 0 0 -1/2 1/2*sqrt(3) 0]~~ |

16 | | ~~[ 0 0 1/2*sqrt(3) 1/2 0]~~ |

17 | | ~~[ 0 0 0 0 -1]~~ |

18 | | ~~}}}~~ |

19 | | ~~The output is a matrix which should really be indexed by conjugacy classes of $S_n$, i.e. partitions. So the order here matters since currently it means that the rows and columns are labelled by~~ |

20 | | ~~{{{~~ |

21 | | ~~sage: Partitions(5).list()~~ |

22 | | ~~[[5], [4, 1], [3, 2], [3, 1, 1], [2, 2, 1], [2, 1, 1, 1], [1, 1, 1, 1, 1]]~~ |

23 | | ~~}}}~~ |

24 | | ~~If the order changes all the matrices need to be changed as well, I think.~~ |

25 | | |