R：使用ROI包

发布于 2025-01-02 19:38:44 字数 952 浏览 0 评论 0原文

我正在尝试使用 R 包 ROI 来解决简单的投资组合优化问题。
我可以使用 Quadprog 求解器“手动”获得结果，但我真的很想了解 ROI 包的工作原理。
不幸的是，我遇到了错误，尽管我坚持使用 Stefan Theussl 在 http://statmath.wu.ac.at/courses/optimization/Presentations/ROI-2011.pdf（幻灯片26,27）

这里是代码：

library(fPortfolio)  
library(ROI)

data(LPP2005.RET)  
lppData <- 100 * LPP2005.RET[, 1:6]
r <- mean(lppData)  
foo <- Q_objective(Q = cov(lppData), L = rep(0, ncol(lppData)))  
full_invest <- L_constraint(rep(1, ncol(lppData)), "==", 1)  
target_return <- L_constraint(apply(lppData, 2, mean), "==",r)
op <- OP(objective = foo, constraints = rbind(full_invest, target_return))   
sol <- ROI_solve(op, solver = "quadprog")

我收到的错误消息是：

(dir == "<=") | 中出现错误(dir = q = "<") : 可以进行操作仅适用于数字、逻辑或复杂类型

感谢您的帮助！

原文

I am trying to use the R package ROI for a simple portfolio optimization problem.
I can get the results using the quadprog solver "manually", but I'd really like to understand how the ROI package works.
Unfortunately I run into an error, even though I am sticking to the provided example by Stefan Theussl at http://statmath.wu.ac.at/courses/optimization/Presentations/ROI-2011.pdf (slide 26,27)

Here is the code:

library(fPortfolio)  
library(ROI)

data(LPP2005.RET)  
lppData <- 100 * LPP2005.RET[, 1:6]
r <- mean(lppData)  
foo <- Q_objective(Q = cov(lppData), L = rep(0, ncol(lppData)))  
full_invest <- L_constraint(rep(1, ncol(lppData)), "==", 1)  
target_return <- L_constraint(apply(lppData, 2, mean), "==",r)
op <- OP(objective = foo, constraints = rbind(full_invest, target_return))   
sol <- ROI_solve(op, solver = "quadprog")

The error message I get is: