声振论坛

 找回密码
 我要加入

QQ登录

只需一步,快速开始

查看: 2571|回复: 0

[其他相关] ORDERPACK 2.0 -- Unconditional, Unique, and Partial Ranking, Sorting, and P

[复制链接]
发表于 2005-5-10 16:42 | 显示全部楼层 |阅读模式

马上注册,结交更多好友,享用更多功能,让你轻松玩转社区。

您需要 登录 才可以下载或查看,没有账号?我要加入

x
<P><a href="http://www.fortran-2000.com/rank/index.html" target="_blank" >http://www.fortran-2000.com/rank/index.html</A></P>

<P><FONT face="Goudy Extra Bold CG ATT, arial, Arial, helvetica" size=5><a href="http://www.fortran-2000.com/rank/index.html" target="_blank" >ORDERPACK 2.0</A><B> Unconditional, Unique, and Partial Ranking, Sorting, and Permutation Fortran 90 source code</B>.</P>
<P>Note that ORDERPACK 2.0 is <a href="http://www.fortran.com/F" target="_blank" >F</A>-compatible.</P>
<P>Illustrative Application</P>
<P></FONT><FONT face="Goudy Extra Bold CG ATT, arial, Arial, helvetica" size=3>In spatial-temporal applications one often wishes to know the nearby observations (say the <EM>m</EM> closest) subject to having these nearby observations also prior in time to the observation itself.  Many of the ways of forming purely spatial neighbors such as those based upon Delaunay triangles do not function well in a spatial-temporal setting. However, partial ranking always works and can function with many more dimensions. Since the number of neighbors, <EM>m</EM>, is much smaller than the number of observations, <EM>n</EM>, the partial ranking algorithm saves a great deal of time relative to full ranking.  For 7,000 observations, the partial algorithm saves almost a factor of 200 in terms of time over the full algorithm. Finding all the neighbors for 100,000 observations takes less than 8 minutes on a 600 Mhz PC, thus making spatial-temporal methods feasible for large data sets on desktop machines. This application was based upon work supported by the National Science Foundation under Grant No. BSC-0136229. Any opinions, findings and conclusions or recommendations expressed in this material are those of the author(s) and do not necessarily reflect the views of the National Science Foundation (NSF).   <a href="http://www.fortran-2000.com/tn1.pdf" target="_blank" >Documentation</A>   <a href="http://www.fortran-2000.com/time_neighbors_files.zip" target="_blank" >Complete Application (source code, pc executable code, data, documentation)</A>  </P></FONT>
回复
分享到:

使用道具 举报

您需要登录后才可以回帖 登录 | 我要加入

本版积分规则

QQ|小黑屋|Archiver|手机版|联系我们|声振论坛

GMT+8, 2024-5-19 07:32 , Processed in 0.128259 second(s), 17 queries , Gzip On.

Powered by Discuz! X3.4

Copyright © 2001-2021, Tencent Cloud.

快速回复 返回顶部 返回列表