File size: 912 Bytes
de46078 | 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 | id: nested_explode_outer_empty
category: schema_nested
difficulty: medium
probes: >
explode() drops rows whose array is empty or null; explode_outer() keeps them
with a NULL element. The fixture contains one empty array and one null array,
so the two functions give different row counts.
tags: [explode, arrays, null_semantics]
prompt: |
Flatten the tags array in `docs` so there is one row per tag, KEEPING documents
that have an empty or null tags array (their tag should be null).
Return columns: doc_id, tag.
fixtures:
- name: docs
schema: doc_id INT, tags ARRAY<STRING>
rows:
- [1, ["x", "y"]]
- [2, []]
- [3, null]
- [4, ["z"]]
solution: |
from pyspark.sql import functions as F
def solve(spark, docs):
# explode() would silently drop docs 2 and 3.
return docs.select("doc_id", F.explode_outer("tags").alias("tag"))
compare:
mode: rows
|