A simple Library for Relation Algebra

• What is Relation Algebra?

  • Relational algebra is a kind of abstract query language. It uses the operation of relation to express the query, as a mathematical tool to study the relation data language. The operation object of relational algebra is relation, and the operation result is relation.

• Why the repo?

  • For skill learning:books:, experimenting:wrench: after class.
  • More for fun!:happy:

Features:fire:

Classic relation operations, 基本关系运算

1. Selection, 选择运算

   • By calling the selection(lambda) method.

   • Define a lambda to choose the rows you want, and use Row.index() to locate a field, just like the follow.

     relation = r.selection(
         lambda row: row[row.index('YOUR_FIELD_NAME')] 
         	== YOUR_FIELD_VALUE
     )

2. Projection, 投影运算

   • Use the projection method.

   • Choose some of the columns you need in the relation object.

     relation = r.projection(['id', 'name'])

3. Extended Cartesian Product, 广义笛卡尔积

   • By using the * operator.

   • The __mul__ method return a Cartesian Product (object of Relation) from two relation operands.

   • Quick and easy, just like the follow.

     relation = r1 * r2

4. Union, 并运算

   • By using the + operator.

   • The __add__ method return a Union from two relation operands.

     relation = r1 + r2

5. Difference, 差运算

   • By using the - operator.

   • The __sub__ method return Difference from two relation operands.

     relation = r1 - r2

Specific relation operations, 特殊关系运算

1. Intersection, 交运算

   • The intersection method operate the self and the other object and return a intersection (object of Relation).

     relation = r1.intersection(r2)

2. Natural Join, 自然连接

   • The "Natural Join" picks the some-named fields and make special connection between two relation objects.

     relation = r1.natural_join(r2)

3. Division, 除运算

   • The most magical operation in all over.

   • By using the operator / .

     relation = r1 / r2

Getting Start:rocket:

• Clone this repo.
• Or, download the zip package.

Then

1. Create a python file in your project directory, or open a Python Console.

2. Import necessary modules.

   >>> from entity.relation import Relation

3. Create an instance of Relation, setting its fields to make init.

   >>> relation = Relation(['id', 'name', 'score'])  # Just like this!

4. Append rows, by calling the add_row() method.

   >>> relation.add_row([1, 'John', 98])  # id: 1, name: John, score: 98
   >>> relation.add_row([2, 'Tom', 80])  # id: 2, name: Tom, score: 80

5. Try your expressions!

   >>> relation.show()  # Print the relation object your just created.
       id       name     score   
       1        John       98    
       2        Tom        80    

   • Define a lambda to make a selection.

     >>> # Select the rows whose 'id' == 2.
     >>> relation.selection(
     >>>		lambda x: x.fields[x.index('id')] == 2
     >>> ).show()
         id       name     score   
         2        Tom        80    

     >>> relation.selection(
     >>> 	lambda x: x.fields[x.index('id')] == 2 
     >>> 	or x.fields[x.index('score')] == 98
     >>> ).show()
         id       name     score   
         1        John       98    
         2        Tom        80    

   • Push your fields in, and get projection return.

     >>> relation.projection(['name', 'score']).show()
        name     score   
        John       98    
        Tom        80    

Quick Start

• You can also open the file src/playground.py in your IDE to get a Quick Start.

ToDo:wrench:

• Some bugs in __mul__, when the method operate the relations with same-name fields.