B*******t
发帖数: 135 | 1 钱多自然有钱多人的办法，比如Goldman给Facebook弄的那个SPV。甚至钱更多人的像
Paulson还可以跟Goldman订制做空房地产的东东。
但我说现在说的是普通人。
我知道有像sharespost这样的private exchange。但是那些private的东西，liquidity
非常差。投资泡沫最重要的一条是liquidity要好，这样到看着差不多了赶紧要抽身走
人的时候不至于拔出来晚了。
我就想问问有没有下面这种vehicle
1) Strongly correlated with the start-up or pre-IPO tech/internet companies
2) Public traded
3) Relatively liquid
目前想到的tech的ETF满足后两条，但是第一条correlation不够强。
com |
|
s***e
发帖数: 267 | 2 Here is one brute force way to calculate for N R.V. with same correlation r:
The correlation matrix must be p.s.d., which is
r*E + (1-r)*I
where E is N*N all 1 matrix and I is N*N identity matrix.
The first one has eigen values (r*N, 0, ..., 0), and the second one has
eigen values (1-r, 1-r, ..., 1-r). Since the second one is identity any set
of orthogonal vector is its eigenvector, so the eigenvalues are summable, i.
e. the eigenvalues for the original cor matrix is:
(1-r+r*N, 1-r, ..., 1-r). T... 阅读全帖 |
|
c**********e
发帖数: 2007 | 3 Thanks. But my problem is to test if two correlations
are equal. For example, proc corr gives corr(x,y)=0.6
and corr(x,z)=0.7. Are the two correlations signicantly
different? (Apparently, sample size makes a difference.) |
|
r****e
发帖数: 3109 | 4 我有12组数据，每一组有32-35人。每组学生考试成绩从高到低排了名次（没有分数，
只有排名）。现在想看另一个变量（比如3分钟内写的作文的长度）和学生的成绩排名
之间的关系。我想应该用spearman's rank correlation,因为成绩只有排名。我的问题
是，我应该12组分开来看，也就是在每组中，都看看correlation是否significant,但
是12组看完后，应该怎么discuss呢？比如说如果在8组中significant,另4组not
significant,etc.应该怎么结论？还是我的分析方法就有问题？
thanks very much in advance for any advice! |
|
l******g
发帖数: 14 | 5 A correlates B, and C correlatsB as well, how can we see A correlated B is
independent with C?
i partial out C in proc corr statement for var A, B. and the r is =0.5 p=0.
08, but i partial out A in proc corr for var C, B, and the r =0.5 p=0.07. so
, how to answer this question?
many thanks. |
|
o******6
发帖数: 538 | 6 ☆─────────────────────────────────────☆
rayleigh (一塌糊涂※生命的坐标) 于 (Wed Feb 13 21:37:49 2008) 提到:
不知道这个版面合适否。。。
我现在有一组数据，从这组数据我能够得到CORRELATION COEFFICIENT，
这个一般的统计书上都有公式，很简单。
我现在需要得到这个CORRELATION COEFFICIENT estimate的variance，
有没有公式可用？
多谢了先。
☆─────────────────────────────────────☆
marioes (back) 于 (Wed Feb 13 21:57:52 2008) 提到:
有，你得找找
☆─────────────────────────────────────☆
rayleigh (一塌糊涂※生命的坐标) 于 (Wed Feb 13 22:18:28 2008) 提到:
。。。能否告诉我哪个paper或者book？ 我google了，没找到。。。
☆─────────────── |
|
s*r
发帖数: 2757 | 7 the p value will only tell you whether the correlation is significantly
different from 0. but the real question may be whether the correlation is
very close to 1
many ways can give you a pvalue. eg the pvalue from the regression
coefficient in a linear regression, and each way will have its problem, eg,
the normality assumption for linear regression if linear regression is used
in a small sample
.
other. |
|
l*****n
发帖数: 72 | 8 We all know correlation will indicate the similarities between two sequences
.
Here we are using the standard Pearson's product-moment coefficient.
For example, if you have
a=[1 -1 -1 -1 -1 ]
b=[1 -1 -1 -1 -1 ]
then obviously, correlation between factor a and b is
corr(a, b)=1;
Let us change b a little bit to c
c=[-1 1 -1 -1 -1 ]
then corr(a, c)=-0.25 < 0, i.e., negative value.
Now, let us attach many many -1 at the end of a and c, i.e.,
a=[1 -1 -1 -1 -1 -1 -1 -1 ...... -1 ]
c=[-1 1 -1 -1 -1 -1 |
|
h*******d
发帖数: 272 | 9 谢谢您的回复
那 放到MINITAB 里 点‘correlation’ 出来CORRELATION MATRIX 后 把
HIGHLY 的去掉 为什么不行呢 我感觉是不行 但就是说不出为什么。。：-（
THANKS AGAIN. |
|
s******g
发帖数: 193 | 10 highly过于subjective了，也许有些data并不highly correlated但是仍然 somewhat
correlated，这对你的分析还是有影响。 |
|
D******n
发帖数: 2836 | 11 u misunderstand correlation.
if R=0.7, it does have a "correlation", it does have a relationship ,statist
ically.
I guess u probably want to say there is no causality relationship between 1
& 2.
but thats not something that can be checked by just looking at ur data. u ne
ed to conduct some control experiments... |
|
b*****n
发帖数: 685 | 12 First of all,
correlation != linear correlation |
|
jl
发帖数: 398 | 13 如果 是 errors ~ t distribution with d.f =2 . auto-correlation 存在吗?
如果不存在, 什么样的 东西 来 measure the auto-correlated structure. 多谢! |
|
s********k
发帖数: 6180 | 14 【 以下文字转载自 EE 讨论区 】
发信人: silverhawk (silverhawk), 信区: EE
标 题: 求助时间序列correlation的问题
发信站: BBS 未名空间站 (Fri May 7 14:43:51 2010, 美东)
假如有两个时间序列A,B. 之间的correlation为rho（假设rho比较大）,有没有什么办
法将A表
示成为一个B和rho的函数？A=f(B,rho).不要求数学意义上相等，数值意义上近似即可
。谢谢 |
|
p********a
发帖数: 5352 | 15 ☆─────────────────────────────────────☆
lyndern (随便) 于 (Wed Mar 3 17:43:44 2010, 美东) 提到:
We all know correlation will indicate the similarities between two sequences
.
Here we are using the standard Pearson's product-moment coefficient.
For example, if you have
a=[1 -1 -1 -1 -1 ]
b=[1 -1 -1 -1 -1 ]
then obviously, correlation between factor a and b is
corr(a, b)=1;
Let us change b a little bit to c
c=[-1 1 -1 -1 -1 ]
then corr(a, c)=-0.25 < 0, i.e., negative value.
Now, let us attach many |
|
s**f
发帖数: 365 | 16 请问如何在SAS里计算：
是survey data，如何用SAS计算两个var的correlation和这个correlation的
confidence interval？
这两个variable可能是：
binary VS binary
binary VS continuous
binary VS ordinal
如果不是survey data的话，用PROC CORR或者PROC FREQ都能算。不过我这个data有
strata，weight，看了半天文献也没找到如何计算confidence interval。
感谢！ |
|
w********u
发帖数: 328 | 17 empirical correlation我在网上查了很久，既查不到具体公式，又查不到例子，我想
解决一个问题，有a,b,c三个变量，每个变量有每天的数值，一共一年，现在想assume
a portfolio that contains non-zero amount of a,b and c. Develop a model that
manage the downside and maximize the profit from the portfolio 用empirical
correlation. 请问，哪位是否可以给一点思路？或者给我一些于此例类似的相关链接
，我找了半天也没有google出来。
万分感谢！！ |
|
s****e
发帖数: 1180 | 18 想要描述不同种类的random variables之间的correlation有可能吗？如random
variable 1是poisson分布的，random variable 2是binomial分布的，他们之间的
correlation有可能用某一statistic描述吗？有没有相关的reference？thank you！ |
|
|
w****a
发帖数: 1623 | 20 我是统计盲，请教大家一个统计问题，如果我有１０００个或者其它数量的大小不同的
金块，我把每个金块称重，量体积，然后把金块重量（Ｙ）和体积（Ｘ）做回归得到，
Ｙ＝ＣＸ，然后有Ｒ＝１，那么重量和体积之间的correlation是spurious
correlation吗？ 谢谢。 |
|
|
T*******I
发帖数: 5138 | 22 一般说来，pearson correlation = linear regression
你的这个问题就是这个一般意义上的pearson correlation or linear regression。 |
|
c*******o
发帖数: 8869 | 23 如果是一对variables, 很容易做correlation, scatter plot 计算r square, 但是如
果是三个变量， 可以做三维的scatter plot, 但是如果计算r square呢？ 还是说只能
分着计算三对pairwise correlation 的r square? |
|
a****g
发帖数: 8131 | 24 有两组数据
group_1 group_2
a a
a a
a a
b
b
b
c
c
c
其中group_1的3个值跟group_2的是correlated
如何做比较呢？
可以想到的是做个2－way anova，
各位大牛有没有更好的方法? thanks |
|
s**5
发帖数: 68 | 25 Hello,被一问题stuck住了,想请教一下各位：
如果有两个Binary random var (X=0 or 1) and (Y=0 or 1),我现在知道他们各自的
marginal probability, Px and Py,并且 X and Y are correlated with correlation
coefficient r, 应该怎么求joint probability Px,y? 先谢谢了！ |
|
p****x
发帖数: 1346 | 26 有一组变量， X1,X2,X3, 通过不同的公式计算出了变量Y1和Y2。
如果我想知道哪种公式更理想， 是否可用regression的方法? 分别看X1,X2,X3和Y1的
correlation, R-Square=0.9, 再看X1,X2,X3和Y2的correlation, R-Square=0.5.
是否可以证明Y1用的公式比Y2更好？
我的顾虑是Y1是由X1,X2,X3计算出来的， 我去看他们的相关性是不是没意义的？
那么这种情况我应该怎么去证明Y1要优于Y2呢？
谢谢！ |
|
b*****e
发帖数: 2511 | 27 如果A和B非线性相关，他们还有correlation吗？比如说 A= B/(1-B),然后直接测A和B
的correlation, 会不会得出一个很小的值？
另外"r2=0.32,P=ns"里的 P=ns 是啥意思？ |
|
G****g
发帖数: 542 | 28 如果A和B非线性相关，他们还有correlation吗？
Yes
比如说 A= B/(1-B),然后直接测A和B
的correlation, 会不会得出一个很小的值？
Might be, but you can try to use nonparametric test, Spearman for example.
另外"r2=0.32,P=ns"里的 P=ns 是啥意思？
non-significant |
|
K****a
发帖数: 67 | 29 请版上好心人帮帮我一下，感激不尽
需要用SAS Monte Carlo simulation，模拟一组实验数据 group randomization trial：
要求outcomes within a group are correlated and binary(0/1)，given
correlation coefficient as rho=0.01.
predictor variables 有categorical (e.g. treatment, gender ....)和continous
类型。
我搜了许多文献，许多写道要用beta-binomial distribution 来生成数据，可是依然
没有弄明白怎么写code,把这个outcome的相关系数rho插进去，或者我怎么确定相关度
是我想要的呢？
谢谢各位
这是网上一个例子：
data test;
>> b0 = -1.3;
>> b1 = 2.2;
>> b2 = 0;
>> s = 0.6;
>> do group=1 to 30;
>> gamma = s*rannor(123... 阅读全帖 |
|
g****n
发帖数: 47 | 30 Correlation分析取数，如果一组数是性别，用0,1代入查相关性也是准的？如果一组数
是医生诊断代码，比如99346,86750,我怎么改成能做correlation分析的数呢?还是只要
是数字就行不用改？ |
|
D******n
发帖数: 2836 | 31 correlation matrix 看的是 predictors之间的关系吧。
如果你问的是binary variable跟continous ariable之间的关系，看哪个measure。
你仍可以看correlation coefficient，然后就可以看KS， AUC。 |
|
l******9
发帖数: 579 | 32 Thanks,
But, in clustering, each data piont is a scalar (a number).
In my problem, each data piont is a vector that contains a group of numbers.
The distance between two vectors is a the linear correlation of teh two
vectors not two data points in the two vectors.
Example,
v1 = [1, 2, 3]
v2 = [4, 6, 8]
v3 = [10, 5, 9]
v1 and v2 has strong correlation than v1 and v3 or v2 and v3.
So,, v1 and v2 should be put in the same cluster but v3 cannot be.
Any help would be appreciated. |
|
e***7
发帖数: 862 | 33 如果我想知道2个变量correlation，都是NUMERIC，一个变量很大，一个很小（from 0
to 1） ，请问怎么求它们的correlation比较valid？ |
|
j*****e
发帖数: 182 | 34 Pearson correlation is not applicable here.
You have miss-interpreted the concept of correlation.
Your data is high-dimensional. Try to specify your population first. Then,
what population parameter are you trying to compare?
The way you compute corr implies that you treat each individual as a
population and genes as the random samples. Is this what you want?
The problem is complex. Please consult an expert on high-dimensional data
analysis in person. |
|
a******1
发帖数: 201 | 35 1. Only professionals can define the variation in your case, the data
analysts can't. Since you are asking the question with the correlation
coefficient, I assume that as a professional in your field (biology?), you
know that is being used to quantize the variation.
2. In your case, you have two samples with size of 45 and 28, and you would
like to compare the means. The classic and best approach is to use t-test as
suggested. But please note that the each sample does not consists of
independent... 阅读全帖 |
|
t**********y
发帖数: 374 | 36 我这里是随机取十个雌性个体 （f1, f2, f3,..., f10), 同样8个雄性个体（m1, m2,.
..,m8），每个个体测了同样10，000个基因的表达值.
雌性个体之间的correlation coefficients计算： 所有组合， (f1,f2), (f1,f3),...
.,(f1,f10),(f2,f3),...,(f2,f10),.....,(f9,f10)
雄性个体同样处理
如前所述，correlation显然不是独立的，所以不敢贸然选test |
|
g**a
发帖数: 2129 | 37 1.what is the platform for your expression data: RNA-seq or expression assay
. For RNA-seq, it is count data. So what ever you guys talked about are
meaningless. Let's assume it is expression assay.
2.Based on assumption in 1, the expression value you were talking about, are
they value/probe or value/gene? If it is value/gene, how did you combine
the reading from different probes within one gene? It is not that easy to
combine the expression from multiple probes in one gene, just like combine
di... 阅读全帖 |
|
a******1
发帖数: 201 | 38 You just don't understand, do you?
From statistics standpoint, LY1, LY2, ... are not sampling individuals (
randomly selected from a population, and independently), so you don't use
them to calculate r between females or males. The OP's problem is to compare
the distributions of (LY1, LY2, ...) (or some characteristics of the
distributions, like one OP called correlation coefficient) between 2 groups:
Female and Male. So that correlation or whatever OP was interested in is
between (LY1, LY2, ...... 阅读全帖 |
|
T*******I
发帖数: 5138 | 39 In my opinion, Pearson's correlation coefficient does work for your case.
You also can try Spearman's correlation coefficient, which is said to be
used with categorically count data. But I believe for your case, both should
be very close to each other. |
|
v*******e
发帖数: 11604 | 40 How do you define "不correlate"? orthogonal? or just P(correlated) < 0.05?
The two are different. |
|
x******6
发帖数: 160 | 41 X is numerator, Y is denominator, and Z is the rate (X/Y). 如果想知道 Z 是不
是 correlate with X， 用Pearson's correlation 可以吗？
谢谢！ |
|
r******g
发帖数: 286 | 42 根据correlation matrix，把Correlation coefficients大的相应predictor挑出来，
然后保留其中一个。除了这个方法外，请教各位高手：还有没有其他方法来处理这个问
题？特别是实际工作中如何处理这个问题? |
|
|
r********n
发帖数: 7441 | 44 【 以下文字转载自 PhotoGear 讨论区 】
发信人: carbonplay (~~ 贪玩), 信区: PhotoGear
标 题: 大家猜，和本版correlation最高的是哪个版啊？
发信站: BBS 未名空间站 (Sun May 6 18:03:40 2012, 美东)
没想到，居然是宝宝版。。。 |
|
b****u
发帖数: 1130 | 45 第一个问题，你把correlation的定义写出来，你就会有眉目了。 |
|
b*******8
发帖数: 37364 | 46 他说的是Correlation，却说成了Causation。女的堕落，不是原因，而是表现之一。不
是女的堕落造成男的堕落，而是社会堕落造成男女一起堕落。
天朝这个大客车，载着男女一起堕落，分不清的俞敏洪，还在挑逗屁民斗屁民，男女互
相骂，是你造成了客车掉河。 |
|
A***a
发帖数: 2211 | 47 喜歡Correlle的,,夠輕,
其他的太重了 |
|
A***a
发帖数: 2211 | 48 //hand
不過
我後來買了stone ware的mug.,,很漂亮,但是microwave手把好燙阿
correle的沒那麼燙 |
|
b********y
发帖数: 5829 | 49 kalman filter 也可以update correlation吧？ |
|
s********k
发帖数: 6180 | 50 应该是normalized cross-correlation。
series |
|
