2 internet doctests failing in findstat.py

[seblabbe]

On Ubuntu 18.04 with 9.3.beta4, Release Date: 2020-12-14 the command

sage -t --optional=sage,internet src/sage/databases/findstat.py

gives

sage -t --warn-long 72.7 --random-seed=0 src/sage/databases/findstat.py
**********************************************************************
File "src/sage/databases/findstat.py", line 118, in sage.databases.findstat
Failed example:
    r = findstat(Permutations, lambda pi: pi.saliances()[0], depth=2); r  # optional -- internet
Expected:
    0: St000476oMp00099oMp00127 (quality [100, 100])
    1: St001497oMp00119oMp00127 with offset 1 (quality [100, 100])
    2: St000147oMp00027oMp00127 (quality [96, 100])
    3: St000054oMp00064oMp00062 with offset 1 (quality [88, 100])
    4: St000740oMp00062 with offset 1 (quality [87, 100])
    ...
Got:
     0: St000740oMp00066 with offset 1 (quality [100, 100])
     1: St000476oMp00099oMp00127 (quality [100, 100])
     2: St001497oMp00119oMp00127 with offset 1 (quality [100, 100])
     3: St000147oMp00027oMp00127 (quality [96, 100])
     4: St000054oMp00066oMp00149 with offset 1 (quality [89, 100])
     5: St000141oMp00025oMp00127 (quality [87, 100])
     6: St001291oMp00127 with offset 1 (quality [87, 86])
     7: St000051oMp00061oMp00069 (quality [87, 86])
     8: St000316oMp00025oMp00127 (quality [87, 86])
     9: St000653oMp00129oMp00127 (quality [87, 86])
    10: St001184oMp00127oMp00064 with offset 1 (quality [87, 86])
    11: St001227oMp00028oMp00127 (quality [87, 86])
    12: St001480oMp00032oMp00127 (quality [87, 86])
    13: St000066oMp00063oMp00149 with offset 1 (quality [54, 86])
    14: St000840oMp00146oMp00127 (quality [18, 86])
    15: St000199oMp00063 with offset 1 (quality [15, 71])
    16: St000193oMp00004oMp00063 with offset 1 (quality [15, 71])
    17: St000200oMp00063oMp00062 with offset 1 (quality [15, 71])
**********************************************************************
File "src/sage/databases/findstat.py", line 129, in sage.databases.findstat
Failed example:
    r[4].info()                                                           # optional -- internet
Expected:
    after adding 1 to every value
    and applying
        Mp00062: Lehmer-code to major-code bijection: Permutations -> Permutations
    to the objects (see `.compound_map()` for details)
    <BLANKLINE>
    your input matches
        St000740: The last entry of a permutation.
    <BLANKLINE>
    among the values you sent, 87 percent are actually in the database,
    among the distinct values you sent, 100 percent are actually in the database
Got:
    after adding 1 to every value
    and applying
        Mp00149: Lehmer code rotation: Permutations -> Permutations
        Mp00066: inverse: Permutations -> Permutations
    to the objects (see `.compound_map()` for details)
    <BLANKLINE>
    your input matches
        St000054: The first entry of the permutation.
    <BLANKLINE>
    among the values you sent, 89 percent are actually in the database,
    among the distinct values you sent, 100 percent are actually in the database
**********************************************************************
1 item had failures:
   2 of  16 in sage.databases.findstat
    [392 tests, 2 failures, 42.52 s]
----------------------------------------------------------------------
sage -t --warn-long 72.7 --random-seed=0 src/sage/databases/findstat.py  # 2 doctests failed
----------------------------------------------------------------------

Upstream: Reported upstream. Developers acknowledge bug.

Component: doctest coverage

Keywords: findstat

Author: Martin Rubey

Branch/Commit: 88d7a84

Reviewer: Sébastien Labbé

Issue created by migration from https://trac.sagemath.org/ticket/31054

[mantepse]

Upstream: Reported upstream. No feedback yet.

[mantepse]

comment:1

It is a bit strange that St000740oMp00062 is not among the hits anymore. This may happen, since St000740oMp00062 and St000740oMp00066 are identical, but should be investigated anyway.

[mantepse]

Branch: u/mantepse/2_internet_doctests_failing_in_findstat_py

[mantepse]

Commit: 913f1b7

[mantepse]

comment:3

The reason for St000740oMp00062 not being among the results anymore is a (known) shortcoming of the findstat search engine: it only keeps only the set of hits with different statistics (ignoring the maps)! The reason for this is that otherwise the result set would often be huge (more than 1000 hits), with most hits not being interesting. Finding a good solution to this problem is being worked on.

---

New commits:

913f1b7

adapt a doctest, cleanup usage of FINDSTAT_MAP_SEPARATOR

[mantepse]

Changed upstream from Reported upstream. No feedback yet. to Reported upstream. Developers acknowledge bug.

[mantepse]

Author: Martin Rubey

[mantepse]

Changed keywords from none to findstat

[sagetrac-git]

Branch pushed to git repo; I updated commit sha1. New commits:

0ff9533

forgot # optional

[sagetrac-git]

Changed commit from 913f1b7 to 0ff9533

[sagetrac-git]

Branch pushed to git repo; I updated commit sha1. New commits:

88d7a84

fix erroneous check whether any maps are applied

[sagetrac-git]

Changed commit from 0ff9533 to 88d7a84

[seblabbe]

comment:8

With the branch on top of beta5, I now get:

Running doctests with ID 2020-12-30-09-59-26-49919bff.
Git branch: 31054
Using --optional=internet,memlimit,sage
Doctesting 1 file.
sage -t --warn-long 72.1 --random-seed=0 src/sage/databases/findstat.py
    [395 tests, 42.75 s]
----------------------------------------------------------------------
All tests passed!
----------------------------------------------------------------------

[seblabbe]

Reviewer: Sébastien Labbé

[vbraun]

Changed branch from u/mantepse/2_internet_doctests_failing_in_findstat_py to 88d7a84