fill(VGAM)
fill()所属R语言包：VGAM
Creates a Matrix of Appropriate Dimension
创建一个适当维数的矩阵
译者：生物统计家园网 机器人LoveR
描述----------Description----------
A support function for the argument xij, it generates a matrix of an appropriate dimension.
的支撑函数的参数xij，它会产生一个合适的尺寸的矩阵。
用法----------Usage----------
fill(x, values = 0, ncolx = ncol(x))
参数----------Arguments----------
参数：x
A vector or matrix which is used to determine the dimension of the answer, in particular, the number of rows. After converting x to a matrix if necessary, the answer is a matrix of values values, of dimension nrow(x) by ncolx.
它被用来确定的答案的尺寸，特别是甲向量或矩阵的行数。转换后x为一个矩阵，如果必要的话，答案是values的值的矩阵，的维度nrow(x)由ncolx。
参数：values
Numeric. The answer contains these values, which are recycled columnwise if necessary, i.e., as matrix(values, ..., byrow=TRUE).
数字。答案包含这些值，如有必要，回收的列，即，作为matrix(values, ..., byrow=TRUE)。
参数：ncolx
The number of columns of the returned matrix. The default is the number of columns of x.
返回的矩阵的列的数量。默认值是数列的x。
Details
详细信息----------Details----------
The xij argument for vglm allows the user to input variables specific to each linear/additive predictor. For example, consider the bivariate logit model where the first/second linear/additive predictor is the logistic regression of the first/second binary response respectively. The third linear/additive predictor is log(OR) = eta3, where OR is the odds ratio. If one has ocular pressure as a covariate in this model then xij is required to handle the ocular pressure for each eye, since these will be different in general. [This contrasts with a variable such as age, the age of the person, which has a common value for both eyes.] In order to input these data into vglm one often finds that functions fill, fill1, etc. are useful.
xij参数vglm允许用户特定于每个线性/添加剂预测的输入变量。例如，考虑到二元Logit模型，第一/秒线性/添加剂的预测中是logistic回归的第一/第二二进制响应。第三个线性/添加剂的预测是log(OR) = eta3，OR的比值比。如果一个人在这个模型中的协变量xij需要处理为每只眼睛的眼压，因为这些东西会有所不同，一般的眼内压。与此相反，一个变量如age，年龄的人，有一个共同的价值的两只眼睛。为了这些数据输入到vglm“”人们经常发现的功能<X >，fill“等都是有益的。
All terms in the xij and formula arguments in vglm must appear in the form2 argument too.
xij和formula参数中的所有条款vglm必须出现在form2参数。
值----------Value----------
matrix(values, nrow=nrow(x), ncol=ncolx), i.e., a matrix consisting of values values, with the number of rows matching x, and the default number of columns is the number of columns of x.
matrix(values, nrow=nrow(x), ncol=ncolx)组成的矩阵中的值，即，values，与匹配x，和默认的列数的行的数目是数列x。
注意----------Note----------
The effect of the xij argument is after other arguments such as exchangeable and zero. Hence xij does not affect constraint matrices.
xij参数的影响后，其他的参数，如exchangeable和zero。因此xij不影响约束矩阵。
Additionally, there are currently 3 other identical fill functions, called fill1, fill2 and fill3; if you need more then assign fill4 = fill5 = fill1 etc. The reason for this is that if more than one fill function is needed then they must be unique. For example, if M=4 then xij = op ~ lop + rop + fill(mop) + fill(mop) would reduce to xij = op ~ lop + rop + fill(mop), whereas xij = op ~ lop + rop + fill1(mop) + fill2(mop) would retain all M terms, which is needed.
此外，还有其他3个相同的fill函数，称为fill1，fill2和fill3，如果你需要更多的分配fill4 = fill5 = fill1等的原因这是，如果一个以上的fill功能是需要的，那么他们必须是唯一的。例如，如果M=4xij = op ~ lop + rop + fill(mop) + fill(mop)将减少至xij = op ~ lop + rop + fill(mop)，，而xij = op ~ lop + rop + fill1(mop) + fill2(mop)将保留所有M条款，这些条款是必要的。
In Examples 1 to 3 below, the xij argument illustrates covariates that are specific to a linear predictor. Here, lop/rop are the ocular pressures of the left/right eye in an artificial dataset, and mop is their mean. Variables leye and reye might be the presence/absence of a particular disease on the LHS/RHS eye respectively.
在下文实施例1~3中，xij参数示出了线性预测特定的协变量。在这里，lop / rop是在人工数据集的左/右眼的眼压力，和mop是他们的平均。变量leye和reye可能会对一个特定的疾病上的左/右轴眼分别的存在/不存在。
In Example 3, the xij argument illustrates fitting the (exchangeable) model where there is a common smooth function of the ocular pressure. One should use regression splines since s in vgam does not handle the xij argument. However, regression splines such as bs and ns need to have the same basis functions here for both functions, and Example 3 illustrates a trick involving a function BS to obtain this, e.g., same knots. Although regression splines create more than a single column per term in the model matrix, fill(BS(lop,rop)) creates the required (same) number of columns.
在实施例3中，xij参数示出拟合模型（可交换的），其中有一个共同的平滑函数的眼压力。每个人都应该使用云形回归以来svgam不处理xij参数。然而，回归样条曲线，如bs和ns需要为这两个函数具有相同的基础功能，和例3所示的功能BS获得，例如，同样的一招节。虽然回归样条曲线创建多模型矩阵中的单个列，每学期，fill(BS(lop,rop))创建所需的（相同的）的列数。
（作者）----------Author(s)----------
T. W. Yee
参考文献----------References----------
http://www.stat.auckland.ac.nz/~yee.
参见----------See Also----------
vglm.control, vglm, multinomial.
vglm.control，vglm，multinomial。
实例----------Examples----------
fill(runif(5))
fill(runif(5), ncol=3)
fill(runif(5), val=1, ncol=3)
# Generate eyes data for the examples below. Eyes are independent (OR=1).[生成眼睛下面的实施例中的数据。眼睛是独立的（OR = 1）。]
nn = 1000 # Number of people[人数]
eyesdat = data.frame(lop = round(runif(nn), 2),
rop = round(runif(nn), 2),
age = round(rnorm(nn, 40, 10)))
eyesdat = transform(eyesdat,
mop = (lop + rop) / 2, # Mean ocular pressure[平均眼压]
op = (lop + rop) / 2, # Value unimportant unless plotting[价值并不重要，除非策划]
# op = lop, # Choose this if plotting[OP =＃选择垂耳兔，如果策划]
eta1 = 0 - 2*lop + 0.04*age, # Linear predictor for left eye[左眼的线性预测]
eta2 = 0 - 2*rop + 0.04*age) # Linear predictor for right eye[线性预测右眼]
eyesdat = transform(eyesdat,
leye = rbinom(nn, size=1, prob=logit(eta1, inverse=TRUE)),
reye = rbinom(nn, size=1, prob=logit(eta2, inverse=TRUE)))
# Example 1[例1]
# All effects are linear[所有的效果是线性的]
fit1 = vglm(cbind(leye,reye) ~ op + age,
family = binom2.or(exchangeable=TRUE, zero=3),
data=eyesdat, trace=TRUE,
xij = list(op ~ lop + rop + fill(lop)),
form2 = ~ op + lop + rop + fill(lop) + age)
head(model.matrix(fit1, type="lm")) # LM model matrix[LM模型矩阵]
head(model.matrix(fit1, type="vlm")) # Big VLM model matrix[大VLM模型矩阵]
coef(fit1)
coef(fit1, matrix=TRUE) # Unchanged with 'xij'[不变“XIJ”]
constraints(fit1)
max(abs(predict(fit1)-predict(fit1, new=eyesdat))) # Predicts correctly[预测正确]
summary(fit1)
## Not run: [＃不运行：]
plotvgam(fit1, se=TRUE) # Wrong, e.g., because it plots against op, not lop.[错了，例如，因为它的图运，不垂耳。]
# So set op=lop in the above for a correct plot.[所以OP =垂耳，在上面正确的图。]
## End(Not run)[＃（不执行）]
# Example 2[例2]
# Model OR as a linear function of mop[模型或拖把的线性函数]
fit2 = vglm(cbind(leye,reye) ~ op + age,
binom2.or(exchangeable=TRUE, zero=NULL),
data=eyesdat, trace=TRUE,
xij = list(op ~ lop + rop + mop),
form2 = ~ op + lop + rop + mop + age)
head(model.matrix(fit2, type="lm")) # LM model matrix[LM模型矩阵]
head(model.matrix(fit2, type="vlm")) # Big VLM model matrix[大VLM模型矩阵]
coef(fit2)
coef(fit2, matrix=TRUE) # Unchanged with 'xij'[不变“XIJ”]
max(abs(predict(fit2)-predict(fit2, new=eyesdat))) # Predicts correctly[预测正确]
summary(fit2)
## Not run: [＃不运行：]
plotvgam(fit2, se=TRUE) # Wrong because it plots against op, not lop.[错误的，因为它图运算，而不是垂耳。]
## End(Not run)[＃（不执行）]
# Example 3. This model uses regression splines on ocular pressure.[实施例3。该模型采用回归样条，对眼部的压力。]
# It uses a trick to ensure common basis functions.[它使用的伎俩，以确保共同的基础功能。]
BS = function(x, ...) bs(c(x,...), df=3)[1:length(x),,drop=FALSE] # trick[招]
fit3 = vglm(cbind(leye,reye) ~ BS(lop,rop) + age,
family = binom2.or(exchangeable=TRUE, zero=3),
data=eyesdat, trace=TRUE,
xij = list(BS(lop,rop) ~ BS(lop,rop) +
BS(rop,lop) +
fill(BS(lop,rop))),
form2 = ~ BS(lop,rop) + BS(rop,lop) + fill(BS(lop,rop)) +
lop + rop + age)
head(model.matrix(fit3, type="lm")) # LM model matrix[LM模型矩阵]
head(model.matrix(fit3, type="vlm")) # Big VLM model matrix[大VLM模型矩阵]
coef(fit3)
coef(fit3, matrix=TRUE)
summary(fit3)
fit3@smart.prediction
max(abs(predict(fit3)-predict(fit3, new=eyesdat))) # Predicts correctly[预测正确]
predict(fit3, new=head(eyesdat)) # Note the 'scalar' OR, i.e., zero=3[注意“标量或，即，零= 3]
max(abs(head(predict(fit3))-predict(fit3, new=head(eyesdat)))) # Should be 0[应为0]
## Not run: [＃不运行：]
plotvgam(fit3, se=TRUE, xlab="lop") # Correct[更正]
## End(Not run)[＃（不执行）]
转载请注明:出自 生物统计家园网(http://www.biostatistic.net)。
注：
注1：为了方便大家学习，本文档为生物统计家园网机器人LoveR翻译而成，仅供个人R语言学习参考使用，生物统计家园保留版权。
注2：由于是机器人自动翻译，难免有不准确之处，使用时仔细对照中、英文内容进行反复理解，可以帮助R语言的学习。
注3：如遇到不准确之处，请在本贴的后面进行回帖，我们会逐渐进行修订。
|
发表于 2012-10-1 15:34:18
