Nov 27 2009
CRU Code & Data A Disaster
It seems the powers that be in the UK wanted CRU to produce a rebuttal by data – except their code and data is messed up they could not make the call to arms:
CRU asked to prepare data for a pre-emptive release in past couple of days but trouble reconciling issues between data bases has stopped this.
Since even CRU cannot reproduce their results, scientifically their results are now legitimately deemed garbage. Everyone who relied on their data for their conclusions has just had the rug pulled out from them by a bunch of PhD quality configuration management on code and data.
QED: start over, get new people to look at the problem.
QED: start over, get new people to look at the problem.
And sack the ones who put you in this position. That would be my suggestion.
As much data as I mangled over the years, I was always reasonably certain that given the original raw data + the code, and I could crank out the final results. It might be annoying and time consuming, but it could be done.
I’m guessing this means they deleted some of their original data sets. Tsk, tsk. Hard to start from scratch when you no longer have ready access to your base data set.
Holy “shades of Haliburton†…
Barack Obama’s climate czar Carol Browner was a board member of one of the leading carbon offset trading companies, APX.
[…] CRU Code & Data A Disaster […]
[…] I mentioned the difference between sound code and ‘PhD level’ code here and here. The best insight into the mess is the HARRY_READ_ME file, which screams out for hundred’s of […]