USHCN Code Released

11 Responses to USHCN Code Released

1. Anthony Watts says:

   July 13, 2012 at 1:15 am

   Nicely done, thank you!

   Reply
   • stevengoddard says:

     July 13, 2012 at 1:17 am

     Enjoy!

     Reply
   • Ben says:

     July 13, 2012 at 12:37 pm

     Ditto, nicely done sir.

     Reply
2. John B., M.D. says:

   July 13, 2012 at 3:03 am

   Thanks for contributing to the scientific process.
   I’ll just sit back and observe.

   Reply
3. Lorne50 says:

   July 13, 2012 at 3:23 am

   nicely done indeed

   Reply
4. Brian G Valentine says:

   July 13, 2012 at 4:21 am

   I used to think, that an indication of GHG warming would be, a decrease of DELTA of (daytime high – night time low) temperatures. Then I realized there is a huge water bias to this, impossible to make a uniform correction to make sense of it. (UHI correction would be nothing compared with something like that).

   Nevertheless I saw some study that did exactly that, and the conclusion was: Demonstration of GHG warming.

   The point is, people can and will do anything at all to demonstrate GHG warming. Meaningful or not, the whole Government supported thing has become, demonstrate GHG warming. It is such a joke, but it is not one bit funny.

   Reply
5. Will Nitschke says:

   July 13, 2012 at 11:46 am

   Imagine your critics actually downloading the data and doing their own analysis. None of them would be bright enough to understand what to do. I suspect downloading music onto their iPod’s is already pushing it.

   Reply
6. Stark Dickflüssig says:

   July 13, 2012 at 4:12 pm

   Nice!

   $ dos2unix Main.cpp USHCN.cpp USHCN.h && clang++ -o USHCN.run Main.cpp USHCN.cpp
   $ file USHCN.run
   USHCN.run: ELF 64-bit LSB executable, x86-64, version 1 (FreeBSD), dynamically linked (uses
   shared libs), for FreeBSD 9.0 (900044), not stripped

   😀

   Reply
   • Stark Dickflüssig says:

     July 13, 2012 at 5:39 pm

     A stripped version built with -O3 -march=k8-sse3 is only 45K or so (-Os gets it down to 40K). The unoptimised version takes ~2.5x as long to run the full US dataset as the -O3 version. Yow!

     Unoptimised, no -march= setting:
378.48 real 354.72 user 6.03 sys
-O3 -march=k8-sse3
137.33 real 125.97 user 5.01 sys
ditto after strip -s
137.70 real 126.41 user 5.05 sys
-Os -march=k8-sse3 (& strip -s, cos it dinnae seem to make any runtime differences)
146.48 real 132.38 user 5.19 sys
     This is all while compiling chromium in the background (it takes a couple of hours & uses 1.8G disk space while building, bleh).

     I don’t have LTO figured out just yet, or I’d fiddle with it, too.

     Reply
7. Bob Tisdale says:

   January 23, 2013 at 8:02 pm

   Steven: And I’ll second or third the Nicely Done.

   For those visitors who aren’t C++/STL literate, please consider providing the Contiguous U.S. temperatures as a single time series. I’d like to compare it to the sea surface temperatures of the U.S. Coastal Waters.

   Reply
8. RobJ says:

   June 26, 2013 at 9:54 am

   Has the govt. quit updating the raw data sets? I can’t find data subsequent to 4/2012.

   Reply