| |
| Reproducible research in cosmology |
|
Reproducibility is at the heart of scientific methodology and experiments must be replicable for a result to be well established.
Necessary condition to truthfully replicate the experiment is that the paper describe it in enough detail that other groups can mimic it.
With the development of computers abd mathemarical methodolgies, algorithms to extract the information from the data are more and more complex, and it becomes amost impossible to reproduce the figure of a given paper. This is especially true in cosmology where the statistical analysis plays a key role. The concept of reproduciple research consists in providing not only the data related to a given paper, but also codes that were used to analyse the data and the scripts that were used for process the data and create the figures. A nice example is the last planck Cosmic Microwave Background (CMB)) results. None of the official Planck PR1 CMB maps are reproducible, while our group has released all informations (data, code, scripts) to reproduce our results.
See all the talk on Reproducible Research I gave at the SAMSI-SAVI Workshop on Astrostatistics, September 21, at Triangle Park, USA in 2012.
|
| |
| Software |
| All packages I have been involved are described at the CosmoStat Software web page. |
|
| |
