import org.apache.spark.sql.SparkSession
import org.apache.spark.sql.Dataset
import org.apache.spark.sql.Row
import org.apache.spark.sql.DataFrame
import org.apache.spark.sql.Column
import org.apache.spark.sql.DataFrameReader
import org.apache.spark.rdd.RDD
import org.apache.spark.sql.catalyst.encoders.ExpressionEncoder
import org.apache.spark.sql.Encoder
import org.apache.spark.sql.functions._
import org.apache.spark.sql.DataFrameStatFunctions
import org.apache.spark.ml.linalg.Vectors
math.sqrt(-1.0)
res43: Double = NaN
math.sqrt(-1.0).isNaN()
res44: Boolean = true
val data1 = data.toDF("affairs", "gender", "age", "yearsmarried", "children", "religiousness", "education", "occupation", "rating")
data1: org.apache.spark.sql.DataFrame = [affairs: string, gender: string ... 7 more fields]
data1.limit(10).show
+-------+------+---+------------+--------+-------------+---------+----------+------+
|affairs|gender|age|yearsmarried|children|religiousness|education|occupation|rating|
+-------+------+---+------------+--------+-------------+---------+----------+------+
| 0| male| 37| 10| no| 3| 18| 7| 4|
| 0| null| 27| null| no| 4| 14| 6| null|
| 0| null| 32| null| yes| 1| 12| 1| null|
| 0| null| 57| null| yes| 5| 18| 6| null|
| 0| null| 22| null| no| 2| 17| 6| null|
| 0| null| 32| null| no| 2| 17| 5| null|
| 0|female| 22| null| no| 2| 12| 1| null|
| 0| male| 57| 15| yes| 2| 14| 4| 4|
| 0|female| 32| 15| yes| 4| 16| 1| 2|
| 0| male| 22| 1.5| no| 4| 14| 4| 5|
+-------+------+---+------------+--------+-------------+---------+----------+------+
// 删除所有列的空值和NaN
val resNull=data1.na.drop()
resNull: org.apache.spark.sql.DataFrame = [affairs: string, gender: string ... 7 more fields]
resNull.limit(10).show()
+-------+------+---+------------+--------+-------------+---------+----------+------+
|affairs|gender|age|yearsmarried|children|religiousness|education|occupation|rating|
+-------+------+---+------------+--------+-------------+---------+----------+------+
| 0| male| 37| 10| no| 3| 18| 7| 4|
| 0| male| 57| 15| yes| 2| 14| 4| 4|
| 0|female| 32| 15| yes| 4| 16| 1| 2|
| 0| male| 22| 1.5| no| 4| 14| 4| 5|
| 0| male| 37| 15| yes| 2| 20| 7| 2|
| 0| male| 27| 4| yes| 4| 18| 6| 4|
| 0| male| 47| 15| yes| 5| 17| 6| 4|
| 0|female| 22| 1.5| no| 2| 17| 5| 4|
| 0|female| 27| 4| no| 4| 14| 5| 4|
| 0|female| 37| 15| yes| 1| 17| 5| 5|
+-------+------+---+------------+--------+-------------+---------+----------+------+
//删除某列的空值和NaN
val res=data1.na.drop(Array("gender","yearsmarried"))
// 删除某列的非空且非NaN的低于10的
data1.na.drop(10,Array("gender","yearsmarried"))
//填充所有空值的列
val res123=data1.na.fill("wangxiao123")
res123: org.apache.spark.sql.DataFrame = [affairs: string, gender: string ... 7 more fields]
res123.limit(10).show()
+-------+-----------+---+------------+--------+-------------+---------+----------+-----------+
|affairs| gender|age|yearsmarried|children|religiousness|education|occupation| rating|
+-------+-----------+---+------------+--------+-------------+---------+----------+-----------+
| 0| male| 37| 10| no| 3| 18| 7| 4|
| 0|wangxiao123| 27| wangxiao123| no| 4| 14| 6|wangxiao123|
| 0|wangxiao123| 32| wangxiao123| yes| 1| 12| 1|wangxiao123|
| 0|wangxiao123| 57| wangxiao123| yes| 5| 18| 6|wangxiao123|
| 0|wangxiao123| 22| wangxiao123| no| 2| 17| 6|wangxiao123|
| 0|wangxiao123| 32| wangxiao123| no| 2| 17| 5|wangxiao123|
| 0| female| 22| wangxiao123| no| 2| 12| 1|wangxiao123|
| 0| male| 57| 15| yes| 2| 14| 4| 4|
| 0| female| 32| 15| yes| 4| 16| 1| 2|
| 0| male| 22| 1.5| no| 4| 14| 4| 5|
+-------+-----------+---+------------+--------+-------------+---------+----------+-----------+
//对指定的列空值填充
val res2=data1.na.fill(value="wangxiao111",cols=Array("gender","yearsmarried") )
res2: org.apache.spark.sql.DataFrame = [affairs: string, gender: string ... 7 more fields]
res2.limit(10).show()
+-------+-----------+---+------------+--------+-------------+---------+----------+------+
|affairs| gender|age|yearsmarried|children|religiousness|education|occupation|rating|
+-------+-----------+---+------------+--------+-------------+---------+----------+------+
| 0| male| 37| 10| no| 3| 18| 7| 4|
| 0|wangxiao111| 27| wangxiao111| no| 4| 14| 6| null|
| 0|wangxiao111| 32| wangxiao111| yes| 1| 12| 1| null|
| 0|wangxiao111| 57| wangxiao111| yes| 5| 18| 6| null|
| 0|wangxiao111| 22| wangxiao111| no| 2| 17| 6| null|
| 0|wangxiao111| 32| wangxiao111| no| 2| 17| 5| null|
| 0| female| 22| wangxiao111| no| 2| 12| 1| null|
| 0| male| 57| 15| yes| 2| 14| 4| 4|
| 0| female| 32| 15| yes| 4| 16| 1| 2|
| 0| male| 22| 1.5| no| 4| 14| 4| 5|
+-------+-----------+---+------------+--------+-------------+---------+----------+------+
val res3=data1.na.fill(Map("gender"->"wangxiao222","yearsmarried"->"wangxiao567") )
res3: org.apache.spark.sql.DataFrame = [affairs: string, gender: string ... 7 more fields]
res3.limit(10).show()
+-------+-----------+---+------------+--------+-------------+---------+----------+------+
|affairs| gender|age|yearsmarried|children|religiousness|education|occupation|rating|
+-------+-----------+---+------------+--------+-------------+---------+----------+------+
| 0| male| 37| 10| no| 3| 18| 7| 4|
| 0|wangxiao222| 27| wangxiao567| no| 4| 14| 6| null|
| 0|wangxiao222| 32| wangxiao567| yes| 1| 12| 1| null|
| 0|wangxiao222| 57| wangxiao567| yes| 5| 18| 6| null|
| 0|wangxiao222| 22| wangxiao567| no| 2| 17| 6| null|
| 0|wangxiao222| 32| wangxiao567| no| 2| 17| 5| null|
| 0| female| 22| wangxiao567| no| 2| 12| 1| null|
| 0| male| 57| 15| yes| 2| 14| 4| 4|
| 0| female| 32| 15| yes| 4| 16| 1| 2|
| 0| male| 22| 1.5| no| 4| 14| 4| 5|
+-------+-----------+---+------------+--------+-------------+---------+----------+------+
//查询空值列
data1.filter("gender is null").select("gender").limit(10).show
+------+
|gender|
+------+
| null|
| null|
| null|
| null|
| null|
+------+
data1.filter("gender is not null").select("gender").limit(10).show
+------+
|gender|
+------+
| male|
|female|
| male|
|female|
| male|
| male|
| male|
| male|
|female|
|female|
+------+
data1.filter( data1("gender").isNull ).select("gender").limit(10).show
+------+
|gender|
+------+
| null|
| null|
| null|
| null|
| null|
+------+
data1.filter("gender<>''").select("gender").limit(10).show
+------+
|gender|
+------+
| male|
|female|
| male|
|female|
| male|
| male|
| male|
| male|
|female|
|female|
+------+