Insyra With Me

What is Insyra

Insyra is an open-source data analysis library for Go. It gives you two core structures, DataList for a column of values and DataTable for a table, plus packages for statistics, charts, machine learning and more, all in plain Go.

Insyra started in August 2024 as a way to do the everyday work of data analysis (load a CSV, clean it, add a column, group, summarize, export) without leaving Go. It has grown into a whole ecosystem since then, but the idea is the same: your analysis is ordinary Go code, so it builds, tests and ships like the rest of your Go project. Insyra is MIT-licensed and built by Hazelnut Paradise.

What it's for

  • Everyday table work. Read CSV, JSON, Excel or Parquet into a DataTable, then filter, sort, group, pivot, merge and write it back out.
  • Excel-style formulas on columns. The Column Calculation Language (CCL) lets you write IF(A > 90, 'Excellent', 'Average') or ['price'] * ['qty'] to build new columns.
  • Statistics you can check. The stats package covers tests, regressions and more, and the project checks its statistics against R.
  • Charts and reports. plot makes interactive charts (go-echarts) and gplot makes static ones (gonum/plot).
  • Models without leaving Go. ml offers scikit-learn-style models and nn runs and trains neural networks in pure Go, with no Python runtime and no cgo.

Install

Insyra v0.3.5 needs Go 1.26 or newer (the go directive in go.mod is 1.26.9):

go get github.com/HazelnutParadise/insyra/allpkgs

A first example

Here is a whole program: put five numbers in a DataList and ask for their mean, median and maximum.

package main

import (
	"fmt"

	"github.com/HazelnutParadise/insyra"
)

func main() {
	scores := insyra.NewDataList(72, 88, 95, 64, 81).SetName("scores")
	fmt.Println("Mean:", scores.Mean())
	fmt.Println("Median:", scores.Median())
	fmt.Println("Max:", scores.Max())
}

Output

Mean: 80
Median: 81
Max: 95

Links