Packages and definitions loaded in the regular kernel are not known by engines and vice versa
- Linguagem predominante
- Jupyter Notebook
- Estrelas
- 2.6k
- Forks
- 1k
- Métricas de merge de PRs
- Nenhum PR com merge em 30d
Descrição
Importing modules or defining functions in regular cells (without the `%%px` magic) are not known in cells where the `%%px` macro is used.
The same is true for the other way around.
To reproduce this try
```py
# Cell 1
import ipyparallel as ipp
from mpi4py import MPI
# Cell 2
cluster = ipp.Cluster(engines="mpi", n=4)
clients = cluster.start_and_connect_sync()
# Cell 3
%%px
import sys
def abc():
print('abc')
# Cell 4
import os
def cde():
print('cde')
# Cell 5:
%%px
os.getpid()
cde()
# Cell 6:
sys.getsizeof(0.0)
abc()
```
Cell 5 and Cell 6 will fail on both statements in each.
A global `%autopx` would ensure that everything runs on the engines, which makes development kind of inconvenient (e.g. prints are done on all engines and so on).
Defining different views and activating them according to how many engines are needed, e.g. one for development with a single engine, seems overkill.
What is the recommended approach to solve the outlined problem?
Guia de contribuição
Direção de pesquisa
Start by reproducing the five-cell example with %%px and the regular kernel, then inspect the documented behavior of %%px and %autopx. The issue does not name implementation files or tests; done would require an agreed approach that makes definitions and imports available across the two execution contexts without requiring global %autopx.
Escrita pelo modelo de indexação a partir do texto da issue.
Avaliação
- Stack de tecnologia
- jupyter, jupyter-notebook, python
- Domínio
- distributed-systems
- Tipo de issue
- Bug
- Dificuldade
- 5/5
- Tempo estimado
- Mais de uma semana
- Status de atividade
- Estagnada
- Clareza
- Precisa de esclarecimento
- Facilidade para iniciantes
- 25/100