Hacktoberfest 2026: as issues que os mantenedores marcaram para outubro, abertas e boas para iniciantes. Ver issues do Hacktoberfest

Restricting a dj.U() aggregation raises a bare AssertionError

Aberta
  • Algo errado?
#1,580 0 comentários 0 reações 0 responsáveis Ver no GitHub

Mantenedores costumam responder em até 1 dia

Ninguém assumiu esta issue ainda.

Avaliação

Dificuldade
3/5
Tempo estimado
1-2 dias
Facilidade para iniciantes
40/100
Tipo de issue
Bug
Clareza
Claramente especificada
Status de atividade
Ativa
Stack de tecnologia
mysql, postgresql, python, sql
Domínio
databases

Direção de pesquisa

The failing assertion is in Aggregation.make_sql in expression.py near line 1358, and the PostgreSQL subquery gate uses _grouping_attributes; read condition.py for the no-common-attributes restriction semantics. First reproduce with dj.U().aggr(File, n="count(*)") & "n = 3" on MySQL. Done when the assertion is gone, the HAVING and subquery paths both return correct rows on MySQL and PostgreSQL, and tests cover string and dict restrictions with matching and non-matching counts plus restricting another table.

Escrita pelo modelo de indexação a partir do texto da issue.

Descrição

Restricting a universal aggregation, dj.U().aggr(...) & <condition>, raises an AssertionError with no message, on any backend. The aggregation itself returns the expected single row; only the restriction fails, including when the restricted aggregate is used to restrict another table.

Versions: v2.3.4 and master at 68fdea59, MySQL 8.0.

Reproduction

@schema
class File(dj.Manual):
    definition = """
    file_id : int32
    ---
    filename : varchar(255)
    """

File.insert([(1, "a_part1.dat"), (2, "a_part2.dat"), (3, "a_meta.json")])
total = dj.U().aggr(File, n="count(*)")

total.to_dicts()                      # [{'n': 3}]
(total & "n = 3").to_dicts()          # AssertionError
(total & {"n": 3}).to_dicts()         # AssertionError
len(File & (total & "n = 3"))         # AssertionError

Expected: (total & "n = 3") returns [{'n': 3}] (and [] when the count differs), and File & (total & "n = 3") keeps every row of File when the restricted aggregate is non-empty and none when it is empty, as restriction by an expression with no common attributes does elsewhere (condition.py, "without common attributes, any non-empty set matches everything").

Cause

Aggregation.make_sql (expression.py:1358):

assert self._grouping_attributes or not self.restriction

A restriction on an aggregation is emitted as HAVING (MySQL) or as an outer WHERE around a subquery (PostgreSQL), but both paths assume grouping attributes exist. The PostgreSQL branch is also gated on them (use_subquery_for_having = ... and self._grouping_attributes).

Proposed fix

  • Remove the assertion.
  • MySQL: emit HAVING without GROUP BY, which MySQL accepts for an ungrouped aggregate and which may reference the select alias (SELECT count(*) AS n FROM t HAVING n = 3).
  • PostgreSQL: take the existing subquery path whenever the aggregation is restricted, with or without grouping attributes.
  • Tests: restrict dj.U().aggr(T, n="count(*)") by a string and by a dict, on both backends, with matching and non-matching counts; and restrict another table by the restricted aggregate.

Workaround

Wrap the aggregate in a subquery before restricting it:

gate = dj.U().aggr(File, n="count(*)").make_subquery() & {"n": 3}
File & gate      # all rows when the count is 3, none otherwise

len(gate) still fails on this form; see the companion issue on len() of an expression with an empty primary key.

Linguagem predominante
Python
Estrelas
197
Forks
98
Merge médio
15h 31min
PRs com merge (30d)
9

Preparar o ambiente

Primeiros passos

  1. Leia a issue inteira e depois o guia de contribuição do projeto.
  2. Comente na issue dizendo que vai assumir — evita que duas pessoas façam o mesmo trabalho.
  3. Faça um fork do repositório e trabalhe em uma branch.
  4. Abra um pull request que referencie o número da issue.

Mais de datajoint/datajoint-python

Todas as issues de datajoint/datajoint-python

Issues semelhantes

Mais issues de Python

Receba novas issues na sua caixa de entrada

Um resumo curto de issues do GitHub para quem está começando.