R读取CSV生成Dataframe后数值列最大值被限制在10以内如何解决

问题描述 #

开展NBA数据可视化个人项目时，读取CSV文件生成R Dataframe后，对得分（PTS）、出手数（FGA）、三分出手数（3PA）等字段做排序和最大值统计时，发现所有列返回的最大值均不超过10：例如斯蒂芬·库里的三分出手数实际大于10，但运行max(stats$3pa)返回结果为9.2，运行max(stats$PTS)返回结果为9.9。

样例数据输出 #

dput(head(stats))的输出如下：

> dput(head(stats))
structure(list(Player = c("Precious Achiuwa", "Steven Adams", 
"Bam Adebayo", "LaMarcus Aldridge", "Ty-Shon Alexander", "Nickeil Alexander-Walker"
), Pos = c("PF", "C", "C", "C", "SG", "SG"), Age = c("21", "27", 
"23", "35", "22", "22"), Tm = c("MIA", "NOP", "MIA", "TOT", "PHO", 
"NOP"), G = c("61", "58", "64", "26", "15", "46"), GS = c("4", 
"58", "64", "23", "0", "13"), MP = c("12.1", "27.7", "33.5", 
"25.9", "3.1", "21.9"), FG = c("2.0", "3.3", "7.1", "5.4", "0.2", 
"4.2"), FGA = c("3.7", "5.3", "12.5", "11.4", "0.8", "10.0"), 
    `FG%` = c(".544", ".614", ".570", ".473", ".250", ".419"), 
    `3P` = c("0.0", "0.0", "0.0", "1.2", "0.1", "1.7"), `3PA` = c("0.0", 
    "0.1", "0.1", "3.1", "0.6", "4.8"), `3P%` = c(".000", ".000", 
    ".250", ".388", ".222", ".347"), `2P` = c("2.0", "3.3", "7.1", 
    "4.2", "0.1", "2.5"), `2PA` = c("3.7", "5.3", "12.4", "8.3", 
    "0.2", "5.2"), `2P%` = c(".546", ".620", ".573", ".505", 
    ".333", ".485"), `eFG%` = c(".544", ".614", ".571", ".525", 
    ".333", ".502"), FT = c("0.9", "1.0", "4.4", "1.6", "0.1", 
    "1.0"), FTA = c("1.8", "2.3", "5.5", "1.8", "0.1", "1.4"), 
    `FT%` = c(".509", ".444", ".799", ".872", ".500", ".727"), 
    ORB = c("1.2", "3.7", "2.2", "0.7", "0.1", "0.3"), DRB = c("2.2", 
    "5.2", "6.7", "3.8", "0.5", "2.8"), TRB = c("3.4", "8.9", 
    "9.0", "4.5", "0.7", "3.1"), AST = c("0.5", "1.9", "5.4", 
    "1.9", "0.4", "2.2"), STL = c("0.3", "0.9", "1.2", "0.4", 
    "0.0", "1.0"), BLK = c("0.5", "0.7", "1.0", "1.1", "0.1", 
    "0.5"), TOV = c("0.7", "1.3", "2.6", "1.0", "0.2", "1.5"), 
    PF = c("1.5", "1.9", "2.3", "1.8", "0.1", "1.9"), PTS = c("5.0", 
    "7.6", "18.7", "13.5", "0.6", "11.0"), c(" ", " ", " ", " ", 
    " ", " ")), row.names = c("1", "3", "4", "5", "6", "7"), class = "data.frame")

根因分析 #

从样例数据可以看到，所有数值列的取值都被引号包裹，说明读取CSV时R将这些字段默认识别为字符型而非数值型。max()函数对字符型数据执行计算时，会按照字符串字典序排序而非数值大小排序，例如字符串"9.9"的字典序大于"18.7"，所以最终返回的最大值均为个位数，和实际数值逻辑不符。

解决方法 #

• 单次计算场景：仅需要临时计算某一列的最大值时，先将列转换为数值型再执行最大值统计，正确写法为：

  max(as.numeric(stats$`3PA`))
  

  注：如果先执行字符型最大值计算再转换数值，仍然会得到错误结果，需注意转换顺序。
• 全局处理场景：如果后续要做大量统计、可视化操作，建议读取CSV时就指定列类型，或者读取后批量将数值列转换为numeric类型，避免后续重复出现同类问题。

内容的提问来源于stack exchange，提问作者CodeMonke