lsddb.org
域名年龄: 13年11个月23天HTTP/1.1 302 Found 访问时间:2016年10月20日 04:03:25 服务器:Apache/2.2.3 (CentOS) 目标网址:http://research.majuric.org/trac/wiki/LargeSurveyDatabase/ 文件大小:316 连接:关闭 类型:text/html; charset=iso-8859-1 HTTP/1.1 302 Found 访问时间:2016年10月20日 04:03:25 服务器:tracd/1.0.1 Python/2.6.8 目标网址:http://research.majuric.org/trac/wiki/LargeSurveyDatabase 类型:text/plain; charset=UTF-8 文件大小:0 其他指令:不缓存 缓存控制:不缓存 过期时间:1999年01月01日 08:00:00 设置Cookie:trac_session=5532a421379676875b7d21d0; expires=Tue, 17-Jan-2017 20:03:25 GMT; httponly; Path=/trac 连接:关闭 HTTP/1.1 200 OK 访问时间:2016年10月20日 04:03:25 服务器:tracd/1.0.1 Python/2.6.8 缓存控制:必须更新 过期时间:1999年01月01日 08:00:00 类型:text/html;charset=utf-8 文件大小:76283 设置Cookie:trac_form_token=0aa24042315d687cefd38d4b; httponly; Path=/trac 设置Cookie:trac_session=cff73d56fc41411fdf05158e; expires=Tue, 17-Jan-2017 20:03:26 GMT; httponly; Path=/trac 连接:关闭 网站编码:utf-8
Search:LoginPreferencesHelp/GuideAbout TracWikiTimelineRoadmapView TicketsSearchwiki:LargeSurveyDatabaseContext NavigationStart PageIndexHistoryLarge Survey DatabaseQuick LinksLarge Survey Database Python API and Referenceschemas of various tablesFrequently Asked QuestionsFuture ConceptsLSD internals for developersAAS 217 LSD posterInstallingwget -N -nv http://lsddb.org/go.sh && bash go.shSee here for more detailed instructions.Sources and Release NotesAlways available on github, but see below for setup instructions and prerequisitesRelease Notes, v0.5.0 (Transactions)Release Notes, v0.5.3 (lsd-import and lsd-admin)Release Notes, v0.5.4 (Remote Database Access)Release Notes, v0.5.5 (User-Defined Functions)Mailing listlsd-users mailing list at Google GroupsIntroductionThe Large Survey Database (LSD) is a framework for storing, cross-matching, querying, and rapidly iterating through large survey datasets (catalogs of >109 rows, >1 TB) on multi-core machines. It is implemented in Python, written and maintained by Mario Juric (mjuric@…).Using LSD can be as simple as writing queries such as:lsd-query --format=fits --bounds='rectangle(180, 10, 190, 20)'\'SELECT ra, dec, cal_psf_mag, filterid, sdss.g as g, sdss.r as r, sdss.i as i FROM ps1_obj, ps1_det, sdss WHERE (0.3 < g-r) & (g-r < 0.4)'and getting the output as a FITS file, or as powerful as writing MapReduce Python kernels. In both cases, the LSD framework will parallelize the workload.The target audience for LSD is anyone who needs to frequently stream through a local copy (of a large chunk thereof) of PS1 catalogs but considers flat text/FITS files too cumbersome, DVO too inflexible (and single-threaded), and a full-fledged RDBMS too slow and/or expensive. Special emphasis is paid to distribution of computation and performance of whole-dataset operations: an LSD (v0.1) setup on a dual Intel Xeon X5560 @ 2.80GHz system (8 physical, 16 logical cores with HT) with 48G of RAM can iterate through 1.1 Grows of PanSTARRS 3Pi data (162GB), cross-matched with 220 Mrows of SDSS data (42G) in ~15 minutes.By ExampleLSD databases are queried using the lsd-query command line utility (described further below), interactively from Python (e.g., from ipython shell), or from Python scripts. In all cases, the query syntax is the same.For example, using lsd-query on the command line, we could write:lsd-query --format=fits --bounds='rectangle(180, 30, 190, 40, coordsys="gal")' 'select ra, dec, r, sdss.r FROM ps1_obj, sdss'This would query the default database (specified via environment variable $LSD_DB) for (ra, dec) and PS1 and SDSS r-band magnitudes of all objects cross-matched between PS1 and SDSS in a Galactic coordinate rectangle bound by l=180, b=30 and l=190, b=40. The output will be stored as a FITS file (named output.fits, by default).If using the Python API, begin with importing the lsd module and instantiating a DB object:from lsd import DBdb = DB('db')where 'db' is the path to th
© 2010 - 2020 网站综合信息查询 同IP网站查询 相关类似网站查询 网站备案查询网站地图 最新查询 最近更新 优秀网站 热门网站 全部网站 同IP查询 备案查询
2025-07-25 09:03, Process in 0.0122 second.