1 回答
TA贡献1828条经验 获得超13个赞
你可以试试下面的代码。
from pyspark.sql.functions import *
# Create DataFrame df1
df1 = spark.createDataFrame([(4,8,1,1), (7,5,6,2), (7,3,5,3)], ["x", "y", "z", "cid"])
# Create DataFrame df2
df2 = spark.createDataFrame([(8,4,5,1), (1,2,9,2), (8,6,4,3), (4,5,4,4)], ["x", "y", "z", "cid"])
# Get the values from cid column from df1
col1 = df1.select(collect_set("cid")).collect()[0][0]
# Filter the dataframe df2 where cid values present in df2 but not in df1.
df3 = df2.filter(~df2["cid"].isin(col1))
# Union df1 and df3.
df4 = df1.union(df3)
df4.show()
// 输出
+---+---+---+---+
| x| y| z|cid|
+---+---+---+---+
| 4| 8| 1| 1|
| 7| 5| 6| 2|
| 7| 3| 5| 3|
| 4| 5| 4| 4|
+---+---+---+---+
我希望这有帮助。
添加回答
举报