R中的二项式GLM:系数谈论什么?
我想知道二项式GLM模型的系数是什么。在我的假设数据中:
# Create the dataset
set.seed(1)
n <- 50
cov <- 10
x <- c(rep(0,n/2), rep(1, n/2))
p <- 0.4 + 0.2*x
y <- rbinom(n, cov, p)
现在,我们使用logit链接:
model0 <- glm(cbind(y, cov-y) ~ x, family="binomial")
summary(model0)
# Call:
# glm(formula = cbind(y, cov - y) ~ x, family = "binomial")
# Deviance Residuals:
# Min 1Q Median 3Q Max
# -1.50013 -0.58688 -0.05123 0.48348 2.43452
# Coefficients:
# Estimate Std. Error z value Pr(>|z|)
# (Intercept) -0.3064 0.1280 -2.394 0.016668 *
# x 0.6786 0.1815 3.739 0.000185 ***
# ---
# Signif. codes: 0 ‘***’ 0.001 ‘**’ 0.01 ‘*’ 0.05 ‘.’ 0.1 ‘ ’ 1
# (Dispersion parameter for binomial family taken to be 1)
# Null deviance: 53.720 on 49 degrees of freedom
# Residual deviance: 39.537 on 48 degrees of freedom
# AIC: 177.99
# Number of Fisher Scoring iterations: 4
返回原始数据量表
model_intercept <- 1/(1+(1/(exp(model0[[1]][1]))))
model_intercept
# (Intercept)
# 0.424
model_x <- 1/(1+(1/(exp(model0[[1]][2]))))
model_x
# x
# 0.6634292
,然后从这里开始,从我开始出现大混乱。 0.6634292值意味着y的增加速率为x的0.6634292。或0.6634292表示使用X间隔使用X间隔为66,34292%的平均增加率。和 关于拦截?尽管X含量为0.4单位,但尽管二项式模型为负值,但Y的含量为0.4单位。
预先感谢您的时间和帮助。
I'd like to uderstand what the coefficents of the binomial GLM models are. In my hypothetical data:
# Create the dataset
set.seed(1)
n <- 50
cov <- 10
x <- c(rep(0,n/2), rep(1, n/2))
p <- 0.4 + 0.2*x
y <- rbinom(n, cov, p)
Now we fit a logistic regression model with x as a covariate, using the logit link:
model0 <- glm(cbind(y, cov-y) ~ x, family="binomial")
summary(model0)
# Call:
# glm(formula = cbind(y, cov - y) ~ x, family = "binomial")
# Deviance Residuals:
# Min 1Q Median 3Q Max
# -1.50013 -0.58688 -0.05123 0.48348 2.43452
# Coefficients:
# Estimate Std. Error z value Pr(>|z|)
# (Intercept) -0.3064 0.1280 -2.394 0.016668 *
# x 0.6786 0.1815 3.739 0.000185 ***
# ---
# Signif. codes: 0 ‘***’ 0.001 ‘**’ 0.01 ‘*’ 0.05 ‘.’ 0.1 ‘ ’ 1
# (Dispersion parameter for binomial family taken to be 1)
# Null deviance: 53.720 on 49 degrees of freedom
# Residual deviance: 39.537 on 48 degrees of freedom
# AIC: 177.99
# Number of Fisher Scoring iterations: 4
Returning to the original data scale
model_intercept <- 1/(1+(1/(exp(model0[[1]][1]))))
model_intercept
# (Intercept)
# 0.424
model_x <- 1/(1+(1/(exp(model0[[1]][2]))))
model_x
# x
# 0.6634292
And here from me starting a big confusion. The 0.6634292 values mean that the y increases at a rate of 0.6634292 per unit of x. Or 0.6634292 means 66,34292% mean increase rate of y with x interval used. And
about the intercept? Something like despite x the y starts at 0.4 units, despite the negative value in the binomial model.
Thanks in advance for your time and help.
如果你对这篇内容有疑问,欢迎到本站社区发帖提问 参与讨论,获取更多帮助,或者扫码二维码加入 Web 技术交流群。

绑定邮箱获取回复消息
由于您还没有绑定你的真实邮箱,如果其他用户或者作者回复了您的评论,将不能在第一时间通知您!
发布评论