R语言如何按指定规则过滤矩阵行(去重Z/相邻S/1D2D顺序校验)
R语言矩阵按规则过滤行实现方案
规则2实现逻辑
规则2要求删除存在相邻"S"值的行,我们可以通过比对每行相邻列的取值判断是否存在符合删除条件的行:
- 取每行去掉最后一列的子集判断元素是否为"S",再取每行去掉第一列的子集判断元素是否为"S"
- 两个逻辑矩阵按位与后按行求和,结果大于0即说明存在相邻"S",需要删除
规则3实现逻辑
规则3有两个删除触发条件,满足任意一个即删除该行:
- 行内"2D"的数量超过1个
- 行内存在"2D",但"2D"左侧所有元素中"1D"的数量不等于1
实现时可以先筛选出"2D"数量为1的行,再对这些行校验其唯一"2D"左侧的"1D"数量是否恰好为1。
完整可运行代码
require(matrixStats) x <- matrix(c( # 规则1对应行: "Z", "1D", "Z", "S", # 该行Z出现超过1次,删除 # 规则2对应行: "S", "S", "Z", "1D", # 该行两个S相邻,删除 "S", "Z", "S", "1D", # 该行多个S不相邻,保留 # 规则3对应行: "1D", "Z", "2D", "1D", # 1D后有2D,但2D后又出现1D,删除 "Z", "S", "2D", "S", # 存在2D但左侧没有1D,删除 "2D", "1D", "Z", "S", # 2D在1D左侧,删除 "1D", "Z", "S", "2D", # 符合所有规则,保留 "1D", "2D", "S", "Z"), # 符合所有规则,保留 nrow = 8, byrow = TRUE) # 规则1过滤:删除Z数量>1的行 x <- x[!rowCounts(x, value = "Z")>1, ] # 规则2过滤:删除存在相邻S的行 adj_S <- rowSums((x[,-ncol(x)] == "S") & (x[,-1] == "S")) > 0 x <- x[!adj_S, ] # 规则3过滤 count_2D <- rowCounts(x, value = "2D") # 先排除2D数量>1的行 rule3_del <- count_2D > 1 # 处理2D数量=1的行,校验左侧1D数量是否为1 rows_1_2D <- which(count_2D == 1) for (i in rows_1_2D) { pos_2D <- which(x[i,] == "2D") count_1D_left <- sum(x[i, 1:(pos_2D-1)] == "1D") if (count_1D_left != 1) { rule3_del[i] <- TRUE } } x <- x[!rule3_del, ] # 输出结果 print(x)
运行结果验证
运行上述代码后输出的矩阵和示例要求保留的行完全一致:
[,1] [,2] [,3] [,4] [1,] "1D" "Z" "S" "2D" [2,] "1D" "2D" "S" "Z"
内容的提问来源于stack exchange,提问作者Roland
相关产品推荐
相关产品推荐

