将规则的先验输出从二进制事务矩阵限制为赋值为 1 的项

Posted 2023-03-23

【中文标题】将规则的先验输出从二进制事务矩阵限制为赋值为 1 的项

【英文标题】：Restrict apriori output of rule from a binary transaction matrix to items assigned value 1

【发布时间】：2018-09-09 16:03:30

【问题描述】：

我正在尝试进行子集化，以使 rhs 和 lhs 上的所有项目都设置为 1，但这不起作用。有没有其他方法可以确保我在规则的 LHS 和 RHS 上的项目都只是“1”？我使用了以下代码。

rules = apriori(cosdat1, parameter=list(support=0.28, confidence=0.3, minlen=2, target="rules"))
summary(rules)

    rules.sub <- subset(rules, subset = lhs %in% c("Bag=1","Blush=1","Nail.Polish=1","Brushes=1","Concealer=1","Eyebrow.Pencils=1","Bronzer=1","Lip.liner=1","Mascara=1","Eye.shadow=1","Foundation=1","Lip.Gloss=1","Lipstick=1","Eyeliner=1"));
    rules.sub

    rules.sub1 <- subset(rules, subset = rhs %in% c("Bag=1","Blush=1","Nail.Polish=1","Brushes=1","Concealer=1","Eyebrow.Pencils=1","Bronzer=1","Lip.liner=1","Mascara=1","Eye.shadow=1","Foundation=1","Lip.Gloss=1","Lipstick=1","Eyeliner=1"));
    rules.sub1

【问题讨论】：

我也试过：nrules

【参考方案1】：

由于您没有提供您的数据，因此有点难以确认，但我相信您已经快到了。您的第一个子集语句给出了 lhs 仅包含 =1 的规则（对 rhs 没有限制）。您的第二个子集语句给出了 rhs 仅包含 =1's 的规则（对 lhs 没有限制）。要使限制适用于双方，请将第二个语句应用于第一个语句的结果。

rules.sub1 <- subset(rules.sub, subset = rhs %in% c("Bag=1","Blush=1",
      "Nail.Polish=1","Brushes=1","Concealer=1","Eyebrow.Pencils=1","Bronzer=1",
      "Lip.liner=1","Mascara=1","Eye.shadow=1","Foundation=1",
      "Lip.Gloss=1","Lipstick=1","Eyeliner=1"));

【讨论】：

有效点。我进行了更改，但我的规则仍然如下所示（items=0）： Bag=0,Brushes=0,Mascara=1 => Eye.shadow=1 Bag=0,Lip.liner= 0,Mascara=1 => Eye.shadow=1 Bag=0,Mascara=1 => Eye.shadow=1 Bag=0,Brushes=0,Eyebrow.Pencils=0,Mascara= 1 => Eye.shadow=1 Nail.Polish=0,Mascara=1 => Eye.shadow=1

当您检查第一个子集时，rules.sub 是否包含 lhs 为零的规则？