Validate rows via join
Another example of filtering data is using joins to remove invalid entries. You'll need to verify the folder names are as expected based on a given DataFrame named valid_folders_df. The DataFrame split_df is as you last left it with a group of split columns.
The spark object is available, and pyspark.sql.functions is imported as F.
Questo esercizio fa parte del corso
Cleaning Data with PySpark
Istruzioni dell'esercizio
- Rename the
_c0column tofolderon thevalid_folders_dfDataFrame. - Count the number of rows in
split_df. - Join the two DataFrames on the folder name, and call the resulting DataFrame
joined_df. Make sure to broadcast the smaller DataFrame. - Check the number of rows remaining in the DataFrame and compare.
esercizio interattivo pratico
Prova questo esercizio completando questo codice di esempio.
# Rename the column in valid_folders_df
valid_folders_df = ____
# Count the number of rows in split_df
split_count = ____
# Join the DataFrames
joined_df = split_df.____(____(valid_folders_df), "folder")
# Compare the number of rows remaining
joined_count = ____
print("Before: %d\nAfter: %d" % (split_count, joined_count))