https://github.com/abhi18av/nextflow-datascience-titanic-survival-analysis

https://github.com/abhi18av/nextflow-datascience-titanic-survival-analysis

data-science nextflow titanic-survival-prediction

Last synced: about 1 month ago
JSON representation

• Host: GitHub
• URL: https://github.com/abhi18av/nextflow-datascience-titanic-survival-analysis
• Owner: abhi18av
• License: mit
• Created: 2020-11-20T09:51:58.000Z (about 4 years ago)
• Default Branch: master
• Last Pushed: 2020-11-27T13:11:18.000Z (about 4 years ago)
• Last Synced: 2023-03-05T19:04:59.112Z (almost 2 years ago)
• Topics: data-science, nextflow, titanic-survival-prediction
• Language: Jupyter Notebook
• Homepage:
• Size: 596 KB
• Stars: 1
• Watchers: 2
• Forks: 0
• Open Issues: 2
• Metadata Files:
  • Readme: README.md
  • License: LICENSE

Awesome Lists containing this project

README

        
Nextflow driven analysis of titanic_dataset

==============================

This project demonstrates how a data-science pipeline can be scaled from a single machine to thousands of machines gracefully. The only thing that needs to be done is 

to wrap the code from Jupyter notebook with Nextflow processes.

### Project structure

```

main.nf

|

modules/

└── data

    ├── test_train_split

    

└── visualization

    ├── gender_survival_plots

    └── survival_plots

|

└── features

    ├── derive_features

    ├── process_age

    ├── process_family

    ├── process_fare

    ├── process_nan

    └── replace_features

    

└── models

    ├── linear_svm

    └── grid_svm

    

workflows/

└── generate_plots

└── feature_engineering

└── train_models

```

## Getting Started

Let's execute the main analysis locally with local data.

```

nextflow run main.nf -entry MAIN -params-file test_params.yml

```

### Overall workflow

The following diagram represents the entire workflow.

![Complete workflow](./workflow.png)

#### Visualization workflow

![Visualization workflow](./docs/visualization_workflow.png)

#### Feature engineering workflow

![Feature Engineering workflow](./docs/feature_engineering_workflow.png)

#### Model training workflow

![Model Training workflow](./docs/model_training_workflow.png)

The foundation for this work is based on  [cookiecutter data science template](https://github.com/drivendata/cookiecutter-data-science) and the 

[Titanic dataset analysis](https://www.kaggle.com/ash316/eda-to-prediction-dietanic)