m*****n 发帖数: 2152 | 1 #include
using namespace std;
int main()
{
double array[13] = {386.34, 250.00, 95.00, 1375.88, 310.11,
1000.00, 1072.96, 333.36, 3179.89, 4339.02, 684.05, 3255.98, 76.71};
int coeff = 0;
for(int coeff=0; coeff < 8192; coeff++)
{
int bit_array[13];
for(int bit_pos=0; bit_pos<13; bit_pos++)
{
if((1 << bit_pos) & coeff) bit_array[bit_pos]=1;
else bit_array[bit_pos]=0;
}
double total =0;
for(int i=0; i |
|
B******a 发帖数: 61 | 2 请问一下Matlab 专家:
coeff = pca(X)
这里coeff 是
PCA= coeff * X
还是 X = coeff * PCA
X is input data |
|
R*****0 发帖数: 146 | 3 没有上下文,不过感觉fit_res$coeffs被默认设成了数组,下标[n]只能用于数组(
矩阵,向量)的运算和读写。试试这个:
fit_res <- list(coeffs = vector("list",100))
for (n in 1:100){
this_fit <- lm(...)
fit_res$coeffs[[n]] <- this_fit$coef
}
fit_res$coeffs[[50]][1:6]
以后直接调用双括号下标[[n]] |
|
c******n 发帖数: 4965 | 4 it's not the absolute value of your coeff,
it's that some intermediate value is 0,
such as the det of your coeff matrix ..... etc,
try change the coeff a bit |
|
s*****n 发帖数: 1279 | 5 Both try1.cpp and try2.cpp can be complied, but only try1.cpp works. when I
run try2.cpp, I always got "Segmentation fault". So what is wrong with try2
.cpp? why I can't use ifstream pointer? Thank you!
try1.cpp
#include
#include
#include
using namespace std;
int main()
{
ifstream gains("coeff.txt");
if (!gains.is_open())
{
cout<<"coeff.txt could not be opened! " <
}
else{
cout<<"coeff.txt is open"<
}
return |
|
c*******n 发帖数: 1648 | 6 Interdiffusion coeff. or selfdiffusion coeff?
If it's inter diffusion coeff. why it's surprising to know it's a negative
number at the phase seperation? If you say self diffusion, then I also want to
know the answer. |
|
K******l 发帖数: 39 | 7 黄金当然是货币,不过目前情况比较复杂,
黄金是=商品属性+被动的金融属性
金融属性之所以被动,是因为劣币依靠国家的强制力量将良币驱逐,将良币变成了所谓
的“商品”, 而这个金融属性波动范围可以大致理解为coeff*max, max是彻底恢复经
典金本位(不是bretton woods那种金汇兑本位)后的黄金价值,coeff取决于劣币的表现
,大家的觉悟,白银或其他东西分享黄金的货币溢价的程度,国家的干预等等。 |
|
J***A 发帖数: 1511 | 8 我说pv= ct 和 p 一定时, 体积正比于温度, 这不都是高中物理热学最基本的东西吗
? 你拿出一个coeff的公式, 能颠覆热力学定律? 你的coeff 反比于温度不也是从体
积正比于温度和他抢不变推出来的?
你和月光呆久了, 最喜欢偷换概念, 比如用cs有sub 考试”来证明工科也有sub 考试
?
不过手抖点回复可能先回你邮箱了....... 见谅见谅。
PV=NRTcoeff=dV/dT/V=NR/PV=1/T你确信你是学工科的? |
|
J***A 发帖数: 1511 | 9 我和chopard 讨论的角度不同, 但我觉得她可能想错了。
因为温度计其实是关心温度变化一度绝对体积的变化。
比如温度从t1 到t2 体积或长度从v1 到v2 求coeff 或者单位体积的变化是无意义的,
因为决定刻度的是总体积, 你求了coeff还得把总体积乘回来 再把温差乘回来, 所以
理想气体温度计, 应该时刻度均匀正比于温度的。
chopard, 大家学高中物理过去这么多年了, 讨论问题时一时想错了都正常, 除非你
每次考试都100分没马虎过......对一个简单问题的讨论, 大家都是求甚解而已,
你第二段是说明 continuous differentiable? |
|
r*****y 发帖数: 53800 | 10 After confusion caused by the UEFA play-offs for 2010 FIFA World Cup
qualifying – which was originally to be an unseeded draw, but was seeded by
FIFA in September 2009 – UEFA have announced that the draw for the play-
offs will be seeded. The four runners-up with the best positions in the UEFA
team coefficient ranking system will be seeded. The draw will define the
four pairings as well as the order of the home and away ties.[6][10] Each
nation's coefficient was generated by calculating:
40% of ... 阅读全帖 |
|
c*******o 发帖数: 62 | 11 求教各位一个算法的实现
用FPGA采集一段信号如图所示,想知道overburst的时间t有多长,
我现在的做法是用min+coeff*(max-min)作阀值来detect transition,
缺点是coeff这个值要不断调试,scdma的信号用0.3,qam的信号用0.1-0.2且效果不好,
不知道有没有更robust的算法能实现 |
|
n***e 发帖数: 11 | 12 我测的abs. coeff.好像和你说的这个不一样。我测的是海水里颗粒物的abs. coeff.。 |
|
a*******r 发帖数: 7558 | 13 abs. coeff.只有从微观颗粒出发才能推出来,无法从宏观理论(RTE)导出来。没有什么
一不一样的,
只有一种定义。
估计你测的是光在叶绿素或CDOM,suspended sediments等颗粒的exintction,好奇你
怎么能分
离出scatt. coeff. |
|
n***e 发帖数: 11 | 14 我测的就是CDOM, phytoplankton, and non-algal particle的 abs. Coeff. 没有
scatt. coeff. 其实很简单,因为我是环境出身,所以对这些太具体的就不知道了。主
要是测个大概的数值,用于remote sensing modeling,还有就是从spectra里分析色素
。 |
|
S********a 发帖数: 359 | 15 说也奇怪,SAS 和R 的coeff estimates and se不同,但是exp以后算成OR 和 OR 的CI
就一样了(或者差别太小忽略不计了?但是PVALUE还是有不一样的),那么这个coeff
不同是由于软件原因,本来就应该有的, 没什么可担心的了? |
|
j**p 发帖数: 53 | 16 would appreciate your help on this on:
I am doing a bunch of lm (~ 100) and wanted to save the results (
coefficients, p value, etc.) nicely to a list
so I have
fit_res = list()
for(n in 1:100)
{
this_fit = lm (....) //I have 5 covariates, so this_fit has 6 coefficients
fit_res$coeffs[n] = this_fit$coefficients
}
however, fit_res$coeffs only contains the first coefficient of each fit, not
the complete set of 6.
thanks! |
|
w****r 发帖数: 28 | 17 fit_res$coeffs[n] 改成
fit_res$coeffs[[n]]
看看 |
|
b********r 发帖数: 764 | 18 为什么不specify好呢?
fit_res$coeffs<-matrix(nrow=100,ncol=6)
fit_res$coeffs[n,]<-as.vector(this_fit$coefficients) |
|
k*****r 发帖数: 21039 | 19 是蛋壳要短胖, 弹头要细长, ballistic coeff 好. |
|
h*******i 发帖数: 4386 | 20 日军的步兵武器,都比较有特点,38大盖子的6.5x50mm子弹,威力适中,子弹偏长,
ballistic coeff很好,外弹道特性突出。作为狙击步枪,非常合适。
50毫米掷弹筒,也是low visual and acoustic signature的近战利器,敌人的机枪通
常还没打完一个弹链,就被干掉了
后期的203毫米火箭炮,就是个63式107毫米火箭炮的放大版,威力巨大,硫磺岛米军伤
亡,主要就是火箭炮干的。 |
|
|
W***n 发帖数: 11530 | 22
When it's hot, since
steel has lower coeff of thermal expansion than aluminum alloy |
|
l***d 发帖数: 1798 | 23 突然想买一个。
walmart.com看来一般。 |
|
C*****d 发帖数: 2253 | 24 PV=NRT
coeff=dV/dT/V
=NR/PV=1/T
你确信你是学工科的? |
|
J***A 发帖数: 1511 | 25 我去查水银和液体温度计的原理了, 为什么刻度是均匀的:) 好像没找到答案, 你
如果找到了请提供链接。 只找到精度很高, 这是不是说明水银的热胀冷缩和理想气体
接近?
想知道温度计的刻度也就是体积或者高度随温度变化的关系 不是应该把体积对温度进
行泰勒展开吗? 所以我觉得温度差就是那个x有问题吗?
温度计刻度的变化不是体积变化delta V吗?
难道是delta v / V?
所以如果水银和理想气体一样, 遵守pv=ct, 难道不是用高中物理就解决了吗.......
为啥要求coeff
你刚才那20分钟是到处去查去搜了吧为啥你的手就这时候抖了? |
|
J***A 发帖数: 1511 | 26 怎么不是呢, 理想气体只是泰勒展开中除了一次项, 其它的高次项系数都是0
如果液体固体和气体不同, 那泰勒就有高次项。
另外如果水银有理想气体的热胀冷缩性质, 那温度变化引起的刻度变化不就是delta V
吗?delta v在一定面积下, 不就是水银柱的高度差吗?
刻度是以某个点为0 的算得是相对的变化, 为何需要算coeff再除以体积这么麻烦呢?
我是没打算颠覆什么定律,问题是你刚才那个贴要是想说的就是“p一定时, 体积正比
于温度”这么件初中生都知道的事,那你提泰勒级数干啥?这么个简单公式是怎么和泰
勒级数扯上关系的?争气........ |
|
J***A 发帖数: 1511 | 27 pv=nrt 和温度计刻度线性问题就是个标准的高考物理题。 根本谈不到大学是工科还是
物理本科的level。这个版上考过大学的都明白。
至少我认为皇上算体积变化的时候normalized by 体积求coeff是把简单题想复杂了。
她最后求温度计刻度变化时会把体积乘回来也说不定。 但如果会乘何必认为我这个一
开始就没除的是错的呢? 所以mit物理系也不能保证高考物理满分。
现在UCLA,UC Berkeley的数学卷子,我拿A+轻松搞定,让皇上回来参加物理高考,
150分满分能拿150, 这个保票,我现在都能给她打。你就不知道物理系的人平时用的
是........ |
|
C*****d 发帖数: 2253 | 28 PV=NRT
coeff=dV/dT/V
=NR/PV=1/T
你确信你是学工科的? |
|
l***n 发帖数: 812 | 29 BDI指数和Drys股价之间关系不大,coeff小于0.2
Drys做长期合同,BDI只是spot market price。
我个人觉得market处于高点,对Drys的pressure很大。另外投资者对Drys进入Tanker的
不乐观,认为该Deleverage的时候却leverage。也有朋友说这是对管理层信息不足造成
的。 |
|
m****0 发帖数: 5857 | 30 用线性回归做了,coeff是406。
除了线性回归还有什么做比较好?
★ 发自iPhone App: ChineseWeb 7.7 |
|
t*******e 发帖数: 216 | 31 it's a fair point.
我想弧度以及弧长同时来控制,
脚下的压力我认为如果在groom过的蓝道上,并且纯carving条件下,应该不会对速度有
很大的控制,因为没有脚下的阻力。
80mph sounds OK. 纯carving,groomed, 应该达到风阻与重力分量平衡的结果。
f = c*v*v = mg*sin(theta), where c is the resistance coeff, theta is the
effective downhill angle (function of curve length, curvature and slope
angle).
风阻系数c与滑雪者的身上衣着、身型以及滑雪姿势有关。所以纯carving平衡速度可以
通过衣着,高低姿势,回转弧度以及弧长来控制。
just for fun, 如果衣着crazy,人其实可以飞起来,那就是遵循贝努利方程了。
如有错误,欢迎指正 |
|
r*****y 发帖数: 53800 | 32 排名是按最近5年的总得分(也就是最后一个coeff)来计算排列的? |
|
g**********y 发帖数: 423 | 33 是这篇:A Genomic Strategy to Elucidate Modules of Oncogenic Pathway
Signaling Networks
另外,我总觉得他们用binreg(svd+probit reg)来预测pathway activity有个问题,就
是他们选的gene signature based on the correlation coeff between phenotype
and the gene exp。如果你仔细看看Bild 2006 的Nature paper,可以看出每个
pathway选出的gene list数目都在变,有时候75,有时候200,300等等,然后再作svd(
效果等同PCA)。这样的feature selection 到底对不对?
就像我们做clustering,只选top highly diff expressed genes做clustering可能有
问题,也可能没事? |
|
l****k 发帖数: 16 | 34 返回的是一个COEFF matix, 看不懂... 根据这个matrix, 怎么知道那些是principle
components? |
|
r****y 发帖数: 1437 | 35 This is up to your input matrix, as mentioned in manual,
row of X is obs, column of x is variable
then coeff(:, 1) is the first PC (in decreasing component variance),
and so on.
scores are then in row manner again, scores(1, :) is the score for
the
first obs vector, and so on.
principle |
|
r**d 发帖数: 21 | 36 it shouldn't be multicollnearity, because if that's the case, R-sq should
increase a lot.
how much R-sq increase after adding a new regressor depends on the partial
correlation coefficient. if i remember correctly, partial correlatin coeff. is
t^2/(t^2+d.o.f.). if the degree of freedom is too high relative to t-value,
the R-sq will not increase much even if t is very large. Greene's book has
several sections on this topic. |
|
n*****1 发帖数: 172 | 37 cluster只管coeff的standard error啊, 跟R^2没有关系吧 |
|
a*******r 发帖数: 7558 | 38 abs. coeff.不就是颗粒的吸收截面乘以颗粒分布密度吗?你是实验室测量还是理论?
Total |
|
c*******n 发帖数: 1648 | 39 Actually what I was talking about is a physical picture in people's mind
When spinodal happens, the tiny amplitude of composition fluctuation wave
exists initally, the composition fluctuation is getting bigger and bigger
with the time. The concentration gradient is getting bigger and bigger as the
result. That's exactly what negative diffusion coeff. means!
When nucleation happens, following your way, let's say tiny amount of
B nucleat into almost 100% pure particles. As this particle grows into |
|
c*******n 发帖数: 1648 | 40 It's interesting, basically you are saying DNA follow Rouse instead Zimm? I
wish I had enough time and could discuss with you intensively and learn sth
for bio guys, but I'd better to disppear in the next two weeks. Otherwise my
boss will be angry. Besides the density, diffusion coeff. will make huge
difference on your result. |
|
c*******n 发帖数: 1648 | 41 You got very good points, Thank you very much. I missed the conceptual
difference between diffusion coeff. and translational energy for the first
question.
Actually, equipartition principle is still widely used in the most common
molecular dynamics software (like insight) to extract temperatures out from
pure momenta, no matter what substance is, asfar as I know. It seems that
you mean the temperature can not be regarded as the single function of the
kinetic energy for liquids, which means the a |
|
w**a 发帖数: 1024 | 42 seems this is a very simple problem.
you can use transformation
t = ln(x)
to reduce the orignal diff. eq. to
y''+b^2 y =g(t) ,here y'' is w.r.t. variable 't'
which is a const. coeff. ode. |
|