My python data analysis container
Architectural configuration
I made a simple DI container with python. Below is a block diagram of the architecture.
About each layer
Application layer
The role of the Application layer is as follows.
- Specify the container creation and logic execution order
- In addition, specify the logic INPUT class and OUTPUT class.
Container layer
The role of the Container layer is as follows.
- DI the logic class, logic INPUT class, and OUTPUT class
- Execute logic in the order specified when creating the container
- Connect logic IN / OUT
Logic layer
Logic execution example)
-
Suppose the following logic is defined.
- SampleLogic1
-
INPUT: None * OUTPU: ISample1Output
- SampleLogic2
- INPUT: ISample1Output
- OUTPU: ISample2Output
- SampleLogic2
-
When creating a container, specify the logic execution order and I / O with the following code.
application.py
logic_dict: LogicDict = LogicDict(
[
{
LogicDict.LOGIC_EXEC_KEY: 'Sample1Logic',
LogicDict.LOGIC_EXEC_INPUT_KEY: '',
LogicDict.LOGIC_EXEC_OUTPUT_KEY: 'ISample1Output',
},
{
LogicDict.LOGIC_EXEC_KEY: 'Sample2Logic',
LogicDict.LOGIC_EXEC_INPUT_KEY: 'ISample1Output',
LogicDict.LOGIC_EXEC_OUTPUT_KEY: 'ISample2Output',
}
]
)
- The logic is DI in the container with
container.execute ()and the logic is executed by the I / O defined in the specified order.
application.py
container = SimulatorContainer(logic_dict)
container.execute()
Usage example
Below is an example of how kaggle can be used for Titanic. https://github.com/chikugoy/kaggle_titanic
Source
The source is published on the following github. https://github.com/chikugoy/analysis_container
Recommended Posts