An individual Python-based service that exposes a simple classifier using scikit-learn and a built-in data-set. The service classifies flower types via distinguishing petal characteristics, petal width, petal length, sepal width, sepal length. It exposes a single function, predict, and demonstrates using system dependencies and data-sets within a module.

type PlantInfo = (Double, Double, Double, Double)
type PlantSpecies = Text
predict : PlantInfo -> PlantSpecies

This service is used to create specific workflows that expose it over http and wire it up to a Postgres database containing sample data to classify.

$ nstack build
$ nstack start Iris.Classify:0.1.0.fromHttp
> Successfully started as process 5
$ nstack send "/iris" '[1.5, 0.1, 4.9, 3.1]'
$ nstack log 5
> ...
> output : "iris-setosa"

$ nstack build
$ nstack start Iris.Classify:0.1.0.fromDb
> Successfully started as process 5
$ nstack log 5