deepbox/dataframe

Series

A one-dimensional labeled array, analogous to a single column of a DataFrame. Supports arithmetic, comparison, and aggregation operations.

new Series

new Series(data: T[], options?: { index?: (string | number)[]; name?: string; copy?: boolean })

Create a Series from an array of values with optional index labels and name.

Parameters:

data: T[] - Array of values for the series

options.name: string - Optional label for the series

options.index: (string | number)[] - Custom index labels (defaults to 0, 1, 2, ...)

Series Methods

.data — Read-only array of values
.name — Series name/label
.length — Number of elements
.index — Index labels
.mean(), .sum(), .min(), .max(), .std(), .var() — Aggregations
.unique() — Unique values
.valueCounts() — Value frequency counts
.sort(ascending?) — Sort values
.map(fn) — Apply function to each element
.filter(fn) — Filter elements by predicate
.toArray() — Convert to plain array
.toTensor() — Convert to Tensor

series.ts

import { Series } from "deepbox/dataframe";const s = new Series([10, 20, 30, 40, 50], { name: "values" });console.log(s.mean());     // 30console.log(s.sum());      // 150console.log(s.std());      // standard deviationconst doubled = s.map((v) => Number(v) * 2);  // [20, 40, 60, 80, 100]const above25 = s.filter((v) => Number(v) > 25); // [30, 40, 50]const t = s.toTensor(); // Convert to Tensor for numerical computing

DataFrame

Descriptive Statistics