You can just group by the yearweek column and then calculate the minimum and maximum dates for each group.

result = df.groupBy("yearweek").agg(
    min("date").alias("first_day_of_week"),
    max("date").alias("last_day_of_week")
)