from pyspark.sql.functions import col, lit
# df is a Spark DataFrame:  DataFrame[basket: string, product: string, customer: string, store: string]
baskets_tally = df.groupBy().agg(countDistinct(col('basket'))).collect()[0][0]
df = df.groupBy(col('basket')).count().withColumnRenamed('count', 'tally_of_products_per_basket')
df = df.groupBy("tally_of_products_per_basket") \
    .count() \
    .withColumnRenamed('count', 'tally_of_baskets_containing_products_tally') \
    .orderBy(col("tally_of_products_per_basket").asc())
df.withColumn(
    'fraction_of_baskets_containing_products_tally',
    col('tally_of_baskets_containing_products_tally') / lit(baskets_tally)
)
### To illustrate the calculated data:
# df.orderBy(col('tally_of_products_per_basket').asc())
#   .select('tally_of_products_per_basket', 'fraction_of_baskets_containing_products_tally').head(10)
#
# returns this histogram:
#
#
#