04 using basic operators - mergen-sergio/DBest GitHub Wiki

The DBest tool provides a wide range of operators that allow users to create complex queries. To get started, focus on the foundational operators of relational algebra, such as:

• Projection
• Selection (Filter)
• Join
• Aggregation

---

The example showcases a query tree that combine basic operators:

• Join: Combines movie and movie_cast tables.
• Filter (Selection): Filters rows based on the year.
• Projection: Projects only the title and character_name columns.

The following screenshots demonstrate how to configure the query tree operators:

1. Join Operator:
   The property window displays the available columns from the left (movie) and right (movie_cast) child nodes. The join predicate is defined as:
   movie.movie_id = movie_cast.movie_id.

   

2. Filter Operator:
   The property window for the filter operator defines an atomic expression that compares the cast_order column to a constant value.

   

3. Projection Operator:
   The property window for the projection operator specifies two columns for retrieval:

   • title from the movie data node.
   • character_name from the movie_cast data node.

   

---

The example above demonstrated basic operators. However, many operators have variations that differ in semantics and efficiency. For instance, the join operation includes:

• Join Types: cross-join, inner-join, semi-join, anti-join, and outer-join.
• Join Algorithms: nested-loop, merge-join, and hash-join.

Start with the classic operators to build a solid foundation. Once comfortable with the basics, experiment with these variations to better understand their use cases and performance implications.

1. Start Small: Use small data sources initially to prevent inefficiencies that may cause the tool to hang.
2. Experiment: Rearrange operators and observe how they affect the result set and query performance.
3. Learn by Doing: Focus on understanding the purpose and behavior of each basic operator before experimenting with advanced operators.