https://github.com/orktes/influunt

Dataflow programming for Golang and Python
https://github.com/orktes/influunt

dataflow-programming golang machine-learning python

Last synced: 9 months ago
JSON representation

Dataflow programming for Golang and Python

• Host: GitHub
• URL: https://github.com/orktes/influunt
• Owner: orktes
• License: mit
• Created: 2018-07-17T06:31:47.000Z (over 7 years ago)
• Default Branch: master
• Last Pushed: 2018-08-12T11:59:42.000Z (over 7 years ago)
• Last Synced: 2024-04-15T12:10:58.495Z (almost 2 years ago)
• Topics: dataflow-programming, golang, machine-learning, python
• Language: Go
• Size: 89.8 KB
• Stars: 5
• Watchers: 2
• Forks: 0
• Open Issues: 0
• Metadata Files:
  • Readme: README.md
  • License: LICENSE

Awesome Lists containing this project

README

          
WORK IN PROGRESS

[![Build Status](https://travis-ci.org/orktes/influunt.svg?branch=master)](https://travis-ci.org/orktes/influunt)

[![GoDoc](https://godoc.org/github.com/orktes/influunt/go?status.svg)](https://godoc.org/github.com/orktes/influunt/go)

# influunt

Dataflow programming for Golang and Python

## Example (create in Python & run in Go)

First create the graph in python and save it to a file

```python

import influunt

with influunt.Graph() as graph:

    # input placeholder

    inputItems = influunt.placeholder()

    # Map items and returns value + 1

    values = inputItems.map(lambda item, i : item.value + 1)

    # Map items and return names

    names = inputItems.map(lambda item, i : item.name)

    # Save graph as a model and define inputs and outputs

    influunt.save_model(

		graph, 

		{"input": inputItems},

		{"values": values, "names": names}, 

		"example.model"

	)

```

Load graph in go and execute

```go

import (

	"fmt"

	"os"

	"github.com/orktes/influunt/go"

	"github.com/orktes/influunt/go/executor"

)

func main() {

	file, _ := os.Open("example.model")

	model, _ := influunt.ReadModel(file)

	exec, _ := executor.NewModelExecutor(model)

	results, _ := exec.Run(map[string]interface{}{

		"input": []map[string]interface{}{

			{"name": "foo", "value": 100},

			{"name": "bar", "value": 200},

		},

	})

	fmt.Printf("%+v\n", results["names"])

	// [foo bar]

	fmt.Printf("%+v\n", results["values"])

	// [101 201]

}

```