Sobes.tech

Data Engineer

What are the disadvantages of UDFs in Spark and what is their downside?

167

Task #3 Find clients with a gap between their orders > 60 days. For these clients, output the maximum gap (gap_days). Orders table: | order_id | customer_id | order_dt | |----------|-------------|------------| | 1 | 1 | [phone] | | 2 | 1 | [phone] | | 3 | 1 | [phone] | | 4 | 1 | [phone] | | 5 | 2 | [phone] | | 6 | 2 | [phone] | | 7 | 3 | [phone] | Final result: | customer_id | gap_days | |-------------|----------| | 1 | 123 | | 2 | 120 |

166

What is the difference between repartition and coalesce in Spark?

166

How would you describe in code a check that a value significantly differs from a normal (expected) distribution?

153

Have you worked with context managers in Python? What are they and what are they used for?

153

If incorrect data has been loaded into the showcase or intermediate table, how can you prevent it from entering higher pipelines?

151

What methods do you know for transferring data between Airflow tasks?

149

What does the depends_on_past parameter do in Airflow?

148

If the DAG schedule is @daily, what time does it actually run?

143

Is a dictionary in Python a mutable or immutable type? What are the requirements for its keys and values?

143

How to run a second Airflow DAG immediately after the first one successfully completes?

133

What is the algorithmic complexity of searching in a dictionary in the worst case?

132
/2