(03-11-05) Reproducibility of Results
(03-11-05) Reproducibility of Results
(03-11-05) Reproducibility of Results
In today's lab meeting, Jasmine mentioned a few points about this topic.- After the submission of a paper, all the data required to reproduce the numbers in the paper should be kept in a directory. Plus a README of how to reproduce them.
- Results must be reproducible by either outside or inside people. If any irreproducibility found, you are ousted.
- To avoid such things to happen, I need to double-check results finally.
- Furthermore, in the middle of the project, I should design pipeline, programs or database tables, etc. in a way easy to check.
Yu Huang 2006-03-25