计算数据帧 Spark 中缺失值的数量 [英] Count the number of missing values in a dataframe Spark

问题描述

我有一个带有缺失值的 dataset ，我想获取每列缺失值的数量.以下是我所做的，我得到了非缺失值的数量.如何使用它来获取缺失值的数量?

I have a dataset with missing values , I would like to get the number of missing values for each columns. Following is what I did , I got the number of non missing values. How can I use it to get the number of missing values?

df.describe().filter($"summary" === "count").show

+-------+---+---+---+
|summary|  x|  y|  z|
+-------+---+---+---+
|  count|  1|  2|  3|
+-------+---+---+---+

如有任何帮助，请获取 dataframe，我们将在其中找到列和每个缺失值的数量.

Any help please to get a dataframe in which we'll find columns and number of missing values for each one.

非常感谢

推荐答案

您可以通过对 isNull() 方法的布尔输出求和，将其转换为整数类型后计算缺失值:

You could count the missing values by summing the boolean output of the isNull() method, after converting it to type integer:

在 Scala 中:

import org.apache.spark.sql.functions.{sum, col}
df.select(df.columns.map(c => sum(col(c).isNull.cast("int")).alias(c)): _*).show

在 Python 中:

from pyspark.sql.functions import col,sum
df.select(*(sum(col(c).isNull().cast("int")).alias(c) for c in df.columns)).show()

或者，您也可以使用 df.describe().filter($"summary" === "count") 的输出，并将每个单元格中的数字减去数据中的行:

Alternatively, you could also use the output of df.describe().filter($"summary" === "count"), and subtract the number in each cell by the number of rows in the data:

在 Scala 中:

import org.apache.spark.sql.functions.lit,

val rows = df.count()
val summary = df.describe().filter($"summary" === "count")
summary.select(df.columns.map(c =>(lit(rows) - col(c)).alias(c)): _*).show

在 Python 中:

from pyspark.sql.functions import lit

rows = df.count()
summary = df.describe().filter(col("summary") == "count")
summary.select(*((lit(rows)-col(c)).alias(c) for c in df.columns)).show()

这篇关于计算数据帧 Spark 中缺失值的数量的文章就介绍到这了，希望我们推荐的答案对大家有所帮助，也希望大家多多支持IT屋！