如何根据一行划分列的所有值？

Question

如何用最后一行的值划分所有行：

col col col3
'A'  2    3
'B'  8    9
'C'  7    5
'fre' 12  13

我想将整个 col2 的值除以 12，将 col3 的值除以 13:

col col col3
'A'  2/12    3/13
'B'  8/12    9/13
'C'  7/12    5/13
'fre' 12/12  13/13

请给出一种对大量列有效的方法。我想对除第一列以外的所有列都这样做。（如上例）

Answer 1

假设最后一列的id是唯一的：

如果您需要遍历大量列

from pyspark.sql.functions import col, when
df = spark.createDataFrame(
    [
        ('A',2,3), 
        ('B',8,9),
        ('C',7,5),
        ('fre',12,13)
    ],
    ['col1','col2','col3']
)

# Get last row
lr = df.tail(1)[0]
# Get last row col1 for otherwise
l_col1 = lr[0]


for c, v in zip(df.columns[1:], lr[1:]):
  df = df.withColumn(c, when(col('col1')!=l_col1, col(c)/v).otherwise(v))

如果最后一列也想分割

from pyspark.sql.functions import col, when
df = spark.createDataFrame(
    [
        ('A',2,3), 
        ('B',8,9),
        ('C',7,5),
        ('fre',12,13)
    ],
    ['col1','col2','col3']
)

# Get last row
lr = df.tail(1)[0]
# Get last row col1 for 
l_col1 = lr[0]


for c, v in zip(df.columns[1:], lr[1:]):
  df = df.withColumn(c, col(c)/v)

如何根据一行划分列的所有值？

how to divide all the values of columns with respect of a row?

pyspark