pyspark中split的简单使用

用于基于某个标识符将字符串/列拆分/断开为多个，并返回列表：

df_b = spark.createDataFrame([('1','ABC-07-DEF')],[ "ID","col1"])
df_b = df_b.withColumn('post_split', F.split(F.col('col1'), "-"))
df_b.show()
+---+----------+--------------+
| ID|      col1|    post_split|
+---+----------+--------------+
|  1|ABC-07-DEF|[ABC, 07, DEF]|
+---+----------+--------------+

此外，还可以使用getitem（）从该arry列中提取列，如下所示

df_b = df_b.withColumn('split_col1', F.col('post_split').getItem(0))
　　.withColumn('split_col2', F.col('post_split').getItem(1))
　　.withColumn('split_col3', F.col('post_split').getItem(2))
df_b.show()
+---+----------+--------------+----------+----------+----------+
| ID|      col1|    post_split|split_col1|split_col2|split_col3|
+---+----------+--------------+----------+----------+----------+
|  1|ABC-07-DEF|[ABC, 07, DEF]|       ABC|        07|       DEF|
+---+----------+--------------+----------+----------+----------+