Removing "." from Spark DataFrame column names

Question

How do I remove the "." from a Spark DataFrame column name?

The DataFrame.select(F.col().alias()) method to rename column names that have a "." in them throws an error.

The following code is reproducible.

# import Spark libraries, configuration, Contexts, and types.
import pyspark
from pyspark.sql import SparkSession
import pyspark.sql.functions as F
#############
# Start Spark.
spark = SparkSession.builder.appName("test").getOrCreate()

testdf = spark.createDataFrame([
    (1, "Julie", "CEO"),
    (2, "Janice", "CFO"),
    (3, "Jake", "CTO")],
    ["ID", "First Name", "Title Initial."])

# this works just fine.
testdf.select(F.col('First Name').alias('first_name')).show(3)

# This throws an error. 
testdf.select(F.col('Title Initial.').alias('title')).show(3)

Error:

AnalysisException: u'syntax error in attribute name: Title Initial.;'

What is an alternative method to change DataFrame column names that have a "." in them?

Answer 1

Surround the column name with `

testdf.select(F.col('`Title Initial.`').alias('title')).show(3)