Question

我正在尝试使用dplyr过滤（）数据帧，我想基于另一个数据框中的列表/值指定的自定义范围进行过滤。

我想要做的一个例子如下。任何帮助将不胜感激。

package week10;

import java.util.*;

/**
 * Skeleton of the recursive implementation of a general tree.
 * 
 * @author Michael Albert
 * @param <T> The type of values stored in the tree.
 */
public class Tree<T> {

    private T rootValue;
    private List<Tree<T>> children;

    public Tree(T rootValue, List<Tree<T>> children) {
        this.rootValue = rootValue;
        this.children = children;
    }

    public Tree(T rootValue) {
        this(rootValue, new ArrayList<Tree<T>>());
    }

    public int size() {
        int count = 1;
        for (Tree<T> child : children) {
            count += child.size();
        }
        return count;
    }

    public int maxDegree() {
        // Get the number of immediate children
        int numChildren = this.children.size();

        // Find the max of all children
        int maxOfChildren = 0;
        for (Tree<T> child : children) {
            maxOfChildren = Math.max(maxOfChildren, child.maxDegree());
        }

        // return the greater of immediate child or max of children
        return Math.max(numChildren, maxOfChildren);
    }

    public void add(Tree<T> child) {
        children.add(child);
    }

    public Tree<T> find(T value) {
        if (rootValue.equals(value)) {
            return this;
        }
        for (Tree<T> child : children) {
            Tree<T> match = child.find(value);
            if (match != null) {
                return match;
            }
        }
        return null;
    }

    public List<T> postOrder() {
        ArrayList<T> list = new ArrayList<T>();
        for (Tree<T> child : children) {
            if (!child.children.isEmpty()) {
                child.postOrder();
            } else {
                //list.add(child);
                System.out.println(child);
            }
        }
        return list;
    }

    public String toString() {
        if (children.isEmpty()) {
            return rootValue.toString();
        }
        return rootValue.toString() + ' ' + children.toString();
    }

    public String toIndentedString() {
        // implement this method
        return "Not implemented yet!";
    }

    /** A helper method for testing (used by main).  Searches tree for
     *  the given target and adds white space separated children to
     *  the tree matching target if there is one.
     *
     * @param target the root value to seach for.
     * @param children a white space separated list of children to add
     * to the tree whose value matches target.
     */
    private static void addChildren(String target, String children) {
        Tree<String> parent = tree.find(target);
        if (parent != null) {
            for (String child : children.split(" ")) {
                parent.add(new Tree<>(child));
            }
        }
    }

    /** A tree instance used for testing. */
    private static Tree<String> tree;

    /**
     * Entry point of the program (used for testing).
     *
     * @param args command line arguments are not used.
     */
    public static void main(String[] args) {
        System.out.println("Creating tree\n-------------");
        tree = new Tree<>("food");
        System.out.print(tree + "\nsize: " + tree.size());
        System.out.println(", max degree: " + tree.maxDegree());
        System.out.println("\nAdding children\n----------------");
        addChildren("food", "meat fruit vegetable");
        System.out.print(tree + "\nsize: " + tree.size());
        System.out.println(", max degree: " + tree.maxDegree());
        System.out.println("\nAdding deeper children\n----------------------");
        addChildren("meat", "chicken beef fish");
        addChildren("fish", "salmon cod tuna shark");
        addChildren("vegetable", "cabbage");
        System.out.print(tree + "\nsize: " + tree.size());
        System.out.println(", max degree: " + tree.maxDegree());
        System.out.println("\nPostorder\n---------");
        System.out.println(tree.postOrder());
        System.out.println("\nIndented string\n---------------");
        System.out.print(tree.toIndentedString());
    }

}

还有另一种方法吗？或者使用现有代码执行此操作？在dplyr包中是否有我遗漏的东西。

非常感谢你。

Answer 1

我认为问题是between期望边界条件的单个值。

您可以尝试：

c<- b %>% filter(source > al & source < ah)

但是，我发现这有点奇怪，因为al和ah将被回收以匹配b$source的大小。

Answer 2

between期望下限/上限的单个值。你可以循环遍历这样的值：

for(i in 1:length(al)){

  print(b %>% filter(between(source,al[i],ah[i])))
}

Answer 3

您可以在dplyr中使用rowwise()执行此操作：

b %>% rowwise() %>% filter(sum((source>al)*(source<ah))>0) %>% ungroup()

返回：

sum((source>al)*(source<ah))返回值下降的间隔数。如果它落在一个区间内，它将返回1.如果它落在2（如果重叠间隔），它将返回2等。

您还可以使用findInterval方法。你可以从al和ah创建一个向量：

v=c(rbind(sort(al),ah[order(al)]))
[1]  0.9  1.1  4.9  5.1 21.9 22.1 35.9 36.1 40.9 41.1

然后查明您的值是否落入奇数或偶数间隔：

b %>% filter(findInterval(source,v)%%2==1)

请注意，此方法仅在您的间隔不重叠时才有效。

在（）之间传递一个列表

3 个答案: