由买买提看人间百态

topics

全部话题 - 话题: censoring
1 2 3 4 5 6 7 8 9 10 下页 末页 (共10页)
G*********o
发帖数: 2045
1
我现在的dataset是censored,但是censor level不是constant,而是每一个
observation有自己的censor level(比如第一个observation是left censor在0,第二
个observation是left censor在10,etc),并且这个censor level是known的。我想在
R里面用tobit model来给这组数据fit一个linear model,但是找了半天只看到有个
VGAM可以做tobit模型,但是只能接受constant censor level,请问有没有人知道其他
package可以用tobit model处理censored data的?包子答谢
l****z
发帖数: 29846
2
Parents: Colorado Teacher Censored Views On Fracking For Middle School
Students
via Breitbart Feed
Student protestors in Jefferson County (JeffCo), Colorado have walked out of
classes this past week, claiming that a reform-minded school board member’
s proposal to review the new A.P. U.S. History curriculum (APUSH) intended
to censor their views. Last year, however, parents accused a teacher at a
JeffCo middle school of presenting only one side of the story on fracking.
As local ABC 7News Denver... 阅读全帖
j*****e
发帖数: 182
3
来自主题: Statistics版 - survival 中关于censoring的问题
You can convert left censored data to right censored date by defining a new
variable, say time_new, such that time_new=l-time_old, where l is a very
large aritifical value. Time_new is then right censored.
As to interval censoring, SAS gives a self-consistent estimator of the
survival function (non-parametric). But that is it, no proportionall hazard
model.
Y******Y
发帖数: 8753
4
来自主题: Statistics版 - 问个censored的survival data概念
censored是没有死或者不知道生死(lost to follow up).
syntax怎么写取决于你怎么define你的censor indicator(假设你叫他ind)
如果你的coding是 ind= 1( if dead), ind=0( if censored)
那syntax就是 model time*ind(0)=var1 var2
如果你的coding是 ind= 0( if dead), ind=1( if censored)
那syntax就是 model time*ind(1)=var1 var2
i****f
发帖数: 432
5
SAS新手,研究了好几天了也还没研究出来,只能试着来求救了。
就是我有一组interval-censored data,其中有一部分是left-censored or right-
censored,其他的都是interval-censored;
作了lifereg以后我怎么才能画出那条regression curve (纵轴是cdf)呢?比如用
lognormal or weibull distribution.
proc lifereg data=xyz;
model (left, right)= / D=weibull;
run;
不知道怎么加图,类似于这篇文章里的那种:
http://www.sciencedirect.com/science/article/pii/S0278691509000
非常感谢!帮忙解决的有包子答谢~
f********8
发帖数: 5601
6
The difference is actually Censoring versus Truncation. Censoring means
corner solutions, the corresponding latent varialbes are negative.
Truncation simply means missing data, no observations are available, so
people put 0 when they collect data. Tobit is suitable for censoring, not
truncation.
j*****e
发帖数: 182
7
Profc reliability gives the turnbull estimator for interval censored/doubled
censored data. This is a pure nonparametric approach. Check the SAS help
withth the key world interval censor or turnbull estimator.
W**********E
发帖数: 242
8
来自主题: Statistics版 - 如何SIMULATE CENSORING PROPROTION?
如何SIMULATE SURVIVAL TIME 并能按预定比例CENSOR(25%,50%,75%)?
比方说TIME TO DEATH ~EXPONENTIAL DIST(RATE=XXX),然后CENSORING TIME~
UNIFORM(0,C)。如何确定这个C最后的数据有预定的CENSORING PROPORTION(25
%,50%,75%)?
谁有这方面的经验
THANKS,
N******n
发帖数: 3003
9
来自主题: Statistics版 - 问个censored的survival data概念
我们检测某种药物对于存活时间的影响, 有很多病人死了,还有活着的,到一定时间
就把所有的人的存活时间记录下来,死了的很容易说,活的就算截止日期,这里是不是
算right censoring, 我的问题是在分析时候,活着的数据是censored, 用一下的SAS程序
proc phreg data=NeuroSur;
model LiveDay*censored(活的) = numberofdrug;
run;
谢谢
r*****t
发帖数: 4793
10
来自主题: Biology版 - Don't censor life-saving science
Don't censor life-saving science
Controlling who is allowed access to information about mutations in the H5N1
bird flu virus is unacceptable, says Peter Palese21.
11 January 2012
Article toolsPrint
Email
Download PDF
Rights and Permissions
Share/bookmark
Connotea
Facebook
Delicious
Twitter
Digg
Friendfeed
The recent arguments over the creation of a transmissible form of the bird
flu virus (H5N1) feel very familiar. My colleagues and I were at the centre
of a similar controversy in 2005, when we ... 阅读全帖
x**0
发帖数: 9
11
有个问题,可能比较基础,不过自己还不是很清楚。希望了解的TX解释一下。
Tobit 与 censored regression model 的区别。
以前觉得前者是后者的一个特例。可是看了两个不同版本的解释,似乎互相矛盾。
wooldridge的教科书里
"While the terms “Tobit” and “censored regression” have often been used
interchangeably in econometrics, in practice there is a very important
difference.The Tobit model is applied to outcome variables that are roughly continuous
over positive values but have a positive probability of equaling zero. We saw an
example of this in the case of married women’s labor supply..
c**********e
发帖数: 2007
12
Use proc lifetime. It is the parametric approach.
proc reliability in SAS/QC is the same as proc lifetime
in SAS/STAT. Be careful for left censored case, one
of them requires left bound as 0 while the other requires
missing (use .).
For interval censored data, SAS only has parametric
approach. For other data, SAS has nonparametric, i.e.,
Kaplan-Meier (proc lifetest), and even more popular
semi-parametric approach. i.e., Cox regression (proc
phreg).
c****u
发帖数: 243
13
来自主题: Statistics版 - 如何SIMULATE CENSORING PROPROTION?
Suppose T is the survival time and C is the censoring time
then censoring rate is Pr(C
j********t
发帖数: 40
14
来自主题: Statistics版 - survival 中关于censoring的问题
在sas中用phreg,sas怎么知道是left censored,right censored还是interval的呢?
还请各位牛人帮忙解惑 :)
j********t
发帖数: 40
15
来自主题: Statistics版 - survival 中关于censoring的问题
谢谢!
那sas中就没有现成的proc可以用来fit semiparametric model for interval
censored or left censored 的数据么? 。。。
j********t
发帖数: 40
16
来自主题: Statistics版 - survival 中关于censoring的问题
没错!!!为什么sas公司的人不开发一个呢?感觉这个实际project中用到interval
或者left censored 的不会没有吧。。。还是因为这种用right censored的方法估计出
来的区别也不大。。。parametric的model有局限性啊。
p****f
发帖数: 251
17
问题应该很常规,就是信用卡公司所能观察到的default都是censored data, 因为那
些打一开始就没发卡的申请人你永远不知道他会不会default。这种情况下,应该怎么
估计default risk?
我的想法是用heckman sample selection correction。但是才意识到以前接触过的
censored data都是针对dependent variable连续的,所以是最普通的linear
regression +selection(probit model),教科书上对于我刚才上面说的那种情况只
字未提。SAS举的例子也全是continous dependent variable+selection的情况。我可
不可以这么写?
proc QLIM data=selectioncorrection;
model binary1= x1 x2/discrete ;
model binary2= x3 x4/discrete select(binary=1);
output out=selectioncrt ; run;
会以包子答谢!
N******n
发帖数: 3003
18
来自主题: Statistics版 - 问个censored的survival data概念

就想问一下这里censored 是 活的, 还是死亡? 因为程序里面censored(1 or 0) 代
表是 alive or death event? In my analysis, I set the alive=1. However I am
not quite sure about that
p*****i
发帖数: 1183
19
Friday, January 22, 2010
(AFP) Google chief executive Eric Schmidt said Thursday that the Internet
giant
is still censoring Web search results in China but that will change in a
"reasonably short time from now."
"We're in conversation with the Chinese government," Schmidt told financial
analysts in a conference call after releasing Google's fourth quarter
results.
"Our business in China is today unchanged," said Schmidt, speaking a little
over
a week after the Internet giant revealed it had come
A*******e
发帖数: 12486
20
【 以下文字转载自 Military 讨论区 】
发信人: Augustine (hehe), 信区: Military
标 题: What Chinese Censors Don’t Want You to Know
发信站: BBS 未名空间站 (Tue Mar 23 23:40:28 2010, 美东)
A set of Chinese government censorship guidelines recently leaked to the
Internet provides a rare and intimate window into the thinking of propaganda
officials. The list of prohibitions issued to editors ranges from the
extremely broad, such as the injunction against “negative news,” to the
bizarrely specific, such as the ban on the blooming
A*******e
发帖数: 12486
21
【 以下文字转载自 Military 讨论区 】
发信人: Augustine (hehe), 信区: Military
标 题: Re: What Chinese Censors Don’t Want You to Know
发信站: BBS 未名空间站 (Tue Mar 23 23:45:25 2010, 美东)
2010年3月6日,值两会召开之际,有网友曝光了3月第一周以来中国网络监管部门向网
站下达的各种监控指令。以下是网络监管部门向某网站下达的“本周网监管理要求”汇
总:
1、两会关于选举法的新闻只能用新华社和人民日报的稿件。
2、关于社会各界要求官员进行财产通报的新闻不报道。
3、国外评南方周末主编为十大风云人物的新闻不报。
4、局长日记的新闻不推荐。新闻中不能出现相关人物图片,或者人肉搜索等涉及个人
隐私的内容。
5、负面信息不上两会要闻区。
6、转载或者发布两会文章时不出现“雷人”、“雷议案”、“雷代表”等字眼,不用
雷的概念定义两会的相关内容。
7、优昙婆罗花的相关新闻删除。
8、海南毒豇豆事件只用新华社、人民日报及海南当地官方媒体的稿件。
9、两会期间关于北京的重
A*******e
发帖数: 12486
22
【 以下文字转载自 Military 讨论区 】
发信人: Augustine (hehe), 信区: Military
标 题: What Chinese Censors Don’t Want You to Know
发信站: BBS 未名空间站 (Tue Mar 23 23:40:28 2010, 美东)
A set of Chinese government censorship guidelines recently leaked to the
Internet provides a rare and intimate window into the thinking of propaganda
officials. The list of prohibitions issued to editors ranges from the
extremely broad, such as the injunction against “negative news,” to the
bizarrely specific, such as the ban on the blooming
A*******e
发帖数: 12486
23
【 以下文字转载自 Military 讨论区 】
发信人: Augustine (hehe), 信区: Military
标 题: Re: What Chinese Censors Don’t Want You to Know
发信站: BBS 未名空间站 (Tue Mar 23 23:45:25 2010, 美东)
2010年3月6日,值两会召开之际,有网友曝光了3月第一周以来中国网络监管部门向网
站下达的各种监控指令。以下是网络监管部门向某网站下达的“本周网监管理要求”汇
总:
1、两会关于选举法的新闻只能用新华社和人民日报的稿件。
2、关于社会各界要求官员进行财产通报的新闻不报道。
3、国外评南方周末主编为十大风云人物的新闻不报。
4、局长日记的新闻不推荐。新闻中不能出现相关人物图片,或者人肉搜索等涉及个人
隐私的内容。
5、负面信息不上两会要闻区。
6、转载或者发布两会文章时不出现“雷人”、“雷议案”、“雷代表”等字眼,不用
雷的概念定义两会的相关内容。
7、优昙婆罗花的相关新闻删除。
8、海南毒豇豆事件只用新华社、人民日报及海南当地官方媒体的稿件。
9、两会期间关于北京的重
k*********t
发帖数: 2999
24
Yesterday, Twitter announced that the micro blogging site will withhold
tweeter posts from a country if they violate any local restrictions rather
than globally blocking them. Till now, twitter had to remove the supposedly
objectionable content entirely from this popular website in order to operate
in such countries.
In a blog post Twitter stated, "As we continue to grow internationally, we
will enter countries that have different ideas about the contours of freedom
of expression."
"Some differ ... 阅读全帖

发帖数: 1
25
GOOGLE’S CEO SAYS TESTS OF CENSORED CHINESE SEARCH ENGINE TURNED OUT GREAT
l****z
发帖数: 29846
26
【 以下文字转载自 GunsAndGears 讨论区 】
发信人: jerseyv (Vendetta), 信区: GunsAndGears
标 题: Obama to Censor Gun Content on Internet
发信站: BBS 未名空间站 (Tue Jun 9 23:56:35 2015, 美东)
j*********r
发帖数: 24733
a***r
发帖数: 594
28
来自主题: USANews版 - suckberg censor这事必须多宣传
this is a classical vietnam war photo, first hand depict of the human
suffering under the brutality of wars. censoring this photo reveals the
contempt and ignorance of facebook as an organization.
b**********5
发帖数: 7881
29
来自主题: JobHunting版 - chinese just love to censor...
An online petition calling for Ellen Pao, the interim CEO of popular website
Reddit, to step down has been signed by over 100,000 people who are deeply
critical of her leadership.
Controversy over Pao's handling of the firing of a popular Reddit
administrator continued to grow, as the uproar prompted several discussion
boards to be taken offline, while the CEO lost more support among Reddit's
users.
According to the Change.org petition, Pao is responsible for "a new age of
censorship" on the web... 阅读全帖
a******r
发帖数: 151
30
来自主题: Parenting版 - leapfrog word builder的censor功能
给女儿新买了一个leapfrog的world builder,发现sex是拼不出来的,多么无谓的
censor功能啊,感叹一下。
c*******9
发帖数: 6411
31
来自主题: Stock版 - weibo censored
this is the key for sina, weibo censored means weibo value will be half of
should-be value, sina much more downside to come.
i might get some if get to $70 level.
o****3
发帖数: 5100
32
来自主题: Soccer版 - Replay needs to be censored
I think what FIFA needs to do: bring two teams in to a black box and
announce the result after 90 minutes. Hell with TV. The games should be
decided by FIFA/Referees, not moder electronics.
http://soccernet.espn.go.com/world-cup/story/_/id/5335440/ce/us/fifa-censor-stadium-replays&cc=5901?ver=us
s********l
发帖数: 245
33
我做的一个project 含有interval censor, 我上网查了一下,如果用SAS的话,要用一
个%ice来实现,可是试了好几次还是不能work,有人有类似的经验吗?能否指导一下,
谢谢!
s********l
发帖数: 245
34
我做的一个project 含有interval censor, 我上网查了一下,如果用SAS的话,要用一
个%ice来实现,可是试了好几次还是不能work,有人有类似的经验吗?能否指导一下,
谢谢!
s********l
发帖数: 245
35
Thanks a lot, I find macro.ice and xmacro online which could use to analysis
interval censor for nonparametric model. But it seems that this procedure
has some requirement on data set.
Thanks again. I gain a lot help from here. I really appreciated!
b**********i
发帖数: 1059
36
谢谢。只发现一个lmec(left censored)
Y******Y
发帖数: 8753
37
来自主题: Statistics版 - survival 中关于censoring的问题
phreg不能accommodate left censoring的情况,lifereg可以,不过fit的不是
semiparametric cox
model...
n******0
发帖数: 298
38
How to add tick marks for censored observations in Kaplan-Meier plot using
survplot in the Design library in R? Thanks.
h******s
发帖数: 3420
39
来自主题: Statistics版 - 问个censored的survival data概念
lost of follow up is also censored
end of follow up date for the patient is not necessarily end day of the
study

程序
b*****h
发帖数: 95
40
Is it about censored regression? Tobit?
Or using conditional MLE since the error is Gaussian.
k*******a
发帖数: 772
41
可以用MLE吧, 假设phi(.)是n(X_i*beta, sigma^2)的CDF f(.)是pdf
如果 y_i=a, likelihood = phi(a)
如果 y_i=b, likelihood = 1 - phi(b)
如果 a < y_i < b, likelihood = f(y_i)
这个和interval censored survival 很类似...
然后optimize这个likelihood
D*********Y
发帖数: 3382
42
比如indicator=1 for mortality event, 0 for censored observation.
model的时候对这俩的处理有什么不同吗?谢谢。
I*****a
发帖数: 5425
43
search survival and censoring a
w*******9
发帖数: 1433
44
你说的是Cox Model. censored data 在partial likelihood 里只起到计算at risk
set 的作用。partial likelihood也是从小白兔说的那个方法来的,就是profile out
the discrete baseline hazard function.
y********o
发帖数: 2565
45
Found the report here:
http://www.npr.org/2012/08/08/158448847/chinas-internet-police-
China's 'Internet Police' Targets Collective Action
August 8, 2012
Audie Cornish talks to Gary King, director of Harvard University's Institute
for Quantitative Social
Science, about new research that looks at the types of online postings
censored by the Chinese
government.
Copyright © 2012 National Public Radio. For personal, noncommercial use
only. See Terms of Use. For other uses, prior
permission requ... 阅读全帖
t***u
发帖数: 20182
46
•Research Article
Reverse-engineering censorship in China: Randomized experimentation and
participant observation
Gary King1,*,
Jennifer Pan1,
Margaret E. Roberts2
+
Author Affiliations
1Institute for Quantitative Social Science, Harvard University, Cambridge,
MA 02138, USA.
2Department of Political Science, University of California, San Diego, La
Jolla, CA 92093, USA.
↵*Corresponding author. E-mail: [email protected]
(function(){try{var s,a,i,j,r,c,l,b=document.getElementsByTagNam... 阅读全帖
w********t
发帖数: 12853
47
Science 22 August 2014:
Vol. 345 no. 6199
DOI: 10.1126/science.1251722
RESEARCH ARTICLE
Reverse-engineering censorship in China: Randomized experimentation and
participant observation
Gary King1,*, Jennifer Pan1, Margaret E. Roberts2
+ Author Affiliations
1Institute for Quantitative Social Science, Harvard University, Cambridge,
MA 02138, USA.
2Department of Political Science, University of California, San Diego, La
Jolla, CA 92093, USA.
↵*Corresponding author. E-mail: [email protected]
(f... 阅读全帖
f**********4
发帖数: 2617
48
这个……我有点费解啊,我不是已经贴出那些章节了么?你手头的罗衰史对照看一下不
就知道吉本花了多少功夫写三世纪危机了吗?莫非你的是节选本?
OK,如果你手头没有,网上也有:http://www.gutenberg.org/files/731/731-h/731-h.htm
我可没有功夫一个皇帝摘一句列出来……OK,就列一段Decius和Valerian好吧,这就超
过几句话了吧:
...At the same time when Decius was struggling with the violence of the
tempest, his mind, calm and deliberate amidst the tumult of war,
investigated the more general causes, that, since the age of the Antonines,
had so impetuously urged the decline of the Roman greatness. He soon
discovered that it was impo... 阅读全帖
c*****1
发帖数: 3240
49
☆─────────────────────────────────────☆
oxthigh (老是肚饿) 于 (Thu Jul 19 12:20:36 2012, 美东) 提到:
欧洲的原始史料可比咱中国的丰富多了,只要你愿意去找
☆─────────────────────────────────────☆
romo (romo) 于 (Fri Jul 20 00:00:15 2012, 美东) 提到:
行为艺术? 欧洲没有老中那样的典藏制度, 古代历史基本靠编。
一个类似的例子是阿三, 阿三的文明程度科比欧洲高多了, 丫的历史还要靠大唐西域
记来还原。 像欧洲那样突然发现柏拉图的理想国几百万字巨著的厚脸皮程度, 棒子是
差太远了。

☆─────────────────────────────────────☆
Diaowai (Waiting for suckers rally) 于 (Fri Jul 20 00:17:59 2012, 美东) 提到:
这个算是闭着眼睛给母猪打口红的例子了。
你读过几部欧洲的“原始资料”?
我是看过几本欧洲古典历... 阅读全帖
y*****w
发帖数: 1350
50
It seems survreg() in R and PROC LIFEREG in SAS run the same type of
survival analysis. However, when I ran both of them on a survival data, I
got different results. Both were set as exponential distribution, and have
right censored data. See below. Could anybody tell me why the results are
different? Did I miss specifying any important parameters in R? Thanks!
~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~
The R code:
survFit <- survreg(Surv(time, event, type="right") ~... 阅读全帖
1 2 3 4 5 6 7 8 9 10 下页 末页 (共10页)