Comments for 10001036

Comment on 转载：公益广告一枚(已更新) by ThinkPad X41和X41T 的一些技巧摘抄

ThinkPad X41和X41T 的一些技巧摘抄 — Thu, 25 Feb 2010 19:44:14 +0000

[...] 与这些software crack相比，下面的hardware crack可能更有意思，图中的三角架是小冰块入学礼物大画板的配件，伸缩自如，长多高都能用。两幅图都挂在Picasa，看不到的冰粉可拜读http://lixiaoxu.lxxm.com/fwadgfw/ [...]

Comment on Wordpress (and WPMU) Plugin for R Web Interface by A Newer Framework « Decision Stats

A Newer Framework « Decision Stats — Sun, 13 Sep 2009 18:27:35 +0000

[...] May 22, 2009 by Ajay Ohri Note this is what I was trying to write in previous post. Also credit is due to http://lixiaoxu.lxxm.com/rwebfriend/ [...]

Comment on Paper for 1st Chinese useR! Conference: Web Powered by R, or R Powered by Web by 10001036 : 《结构方程模型及其应用》(侯, 温, & 成,2004)部分章节R代码

10001036 : 《结构方程模型及其应用》(侯, 温, & 成,2004)部分章节R代码 — Tue, 07 Jul 2009 18:08:38 +0000

[...] [update]R中的sem包妙处之一是可在线实现结构方程应用界面。下面这个最粗陋的例子对应原书Chap3_1_4_CFA_MB.LS8的结果。 [...]

Comment on [二度更新并推荐]愉快地发现SciTE和LyX在WinXP下都支持中文 by joomlagate.com

joomlagate.com — Sun, 26 Apr 2009 09:02:19 +0000

您好！

我今天刚刚安装了 LyX 1.6.2 版本，发现输出 PDF 时中文不能正常输出。看了您的文章，我去下载南开大学的中文插件，发现页面打不开了。

您能否提供一下该插件的下载地址？（最好发到我的 email 信箱）

谢谢！

Comment on My first wp plugin work: LaTeX_Math_cgi 1.0 by lixiaoxu

lixiaoxu — Thu, 02 Apr 2009 20:35:33 +0000

比较痛苦的解决方法是自己提供修改源代码的cgi服务端，为\提供替代符号

Comment on 气泡图击败Data Snoop by 魏太云

魏太云 — Thu, 26 Mar 2009 03:24:39 +0000

这个图我也画过，还用聚类的方法划分过地震带。

Comment on My first wp plugin work: LaTeX_Math_cgi 1.0 by 谢益辉

谢益辉 — Fri, 20 Mar 2009 07:31:04 +0000

请教个问题啊，LaTeX公式中的"\"（backslash）在文章第二次编辑的时候会丢失，这个问题咋个解决啊？Google了一下有解决方案说修改核心PHP程序的，但我觉得最好不要改核心程序，要不然下次升级还得重新改……

Comment on Why practitioners discretize their continuous data by 谢益辉: 统计学博文导读：火箭队比赛与分类树、神经网络与降维 | 统计之都

谢益辉: 统计学博文导读：火箭队比赛与分类树、神经网络与降维 | 统计之都 — Sun, 15 Mar 2009 10:04:34 +0000

[...] 李晓煦老师的博客：非常专业，为数不多的会用LaTeX写上数学公式的博客，李老师对统计理论细节研究很认真，很有国外统计研究者的风范；博文如Why practitioners discretize their continuous data讲述了为什么大家喜欢将连续型数据离散化的原因之一。 [...]

Comment on Why practitioners discretize their continuous data by lixiaoxu

lixiaoxu — Sun, 08 Mar 2009 09:56:27 +0000

Residuals and errors are different. The more intervals, squared-residuals decrease while squared-errors increase. So the black points, or discretization with max intervals, predict red population the worst.

Discretization fades micro information (most errors) while highlights macro information (usually non-linear). When LOESS is popular enough, discretization will be abandoned. Practitioners really need local smoothing to preview their concerned macro models.

Comment on Why practitioners discretize their continuous data by Yihui

Yihui — Sun, 08 Mar 2009 07:26:08 +0000

The discretization here is essentially a kind of local smoothing techniques using a constant kernel function. Generally speaking, local modeling can effectively improve fitness (lower error sum of squares) but we have to carefully avoid overfitting. If you discretize x into more intervals, the fitting will be even better.