Recently, I studied and listened to the course “reproducible research” on Coursera under the data science specialization. At the very beginning, I am little confused about the importance of reproducible research, since generally for us ( financial engineer), it seems unnecessary to reproduce other’s results. The version control, Rpub, cacher package, usage of knir and sweave, all of those issues seems meaningless for financial data analysis. (Honestly, it is. But learning some knowledge is not a bad thing.)
However, today, just read the news about Chunyu Han, say, Zhouzi Fang arises his doubt on the astounding paper Han has just published. And lots of researchers are all focusing on trying to reproduce the Fig 4 in Han’s paper.
Combined those two things, I was just wondering why reproducible research is such important especially in biology area, and I found the answer from the presentation given by Keith A. Baggerly, who is the professor of Bioinformatics and Computational Biology, UT M.D. Anderson Cancer Center.
Here is the link to his presentation video.