df-script

v2.3.0

Published

21 days ago

A zero-dependency, high-performance, expression-based DataFrame library for TypeScript/JavaScript.

🚀 df-script: High-Performance TypeScript DataFrame Library

df-script is a blazing-fast, zero-dependency, expression-based DataFrame and data manipulation library for TypeScript and JavaScript.

Heavily inspired by modern columnar engines like Polars and Pandas, df-script brings declarative, high-performance data transformations directly to JavaScript runtimes (Node.js, Browser, Bun, Deno, and Edge Workers). Easily execute complex multi-column groupings, relational joins, windowed rolling statistics, string parsing, temporal conversions, nested structs, and hundreds more with a fluent, expressive API and zero external dependencies.

📦 Installation

npm install df-script

Or with Yarn, PNPM, or Bun:

yarn add df-script
pnpm add df-script
bun add df-script

Import in TypeScript or JavaScript:

// ES Module / TypeScript
import { $df, DataFrame, ColumnExpr, DataType } from "df-script";

// CommonJS / Node.js
const { $df, DataFrame, ColumnExpr, DataType } = require("df-script");

🚀 Quick Start

1. Basic DataFrame & Expression Pipeline

import { $df } from "df-script";

// Create a DataFrame with structured data and automatic schema inference
const df = $df.data([
  { id: 1, name: "Alice", joinDate: "2026-01-15", sales: 1200.50, tags: ["sales", "east"] },
  { id: 2, name: "Bob", joinDate: "2026-02-20", sales: 850.00, tags: ["support", "west"] },
  { id: 3, name: "Charlie", joinDate: "2026-03-05", sales: 2300.00, tags: ["sales", "north"] },
  { id: 4, name: "David", joinDate: "2026-03-12", sales: null, tags: ["marketing"] },
]);

// Select columns, transform strings, format dates, and compute expressions
const processedDf = df.select(
  $df.col("id"),
  $df.col("name").str.upper().alias("NAME_UPPER"),
  $df.col("joinDate").str.toDatetime().dt.year().alias("joinYear"),
  $df.col("sales").add(500).alias("salesAdjusted"),
  $df.col("tags").arr.lengths().alias("tagCount")
);

console.log(processedDf.toDicts());
/* Output:
[
  { id: 1, NAME_UPPER: 'ALICE', joinYear: 2026, salesAdjusted: 1700.5, tagCount: 2 },
  { id: 2, NAME_UPPER: 'BOB', joinYear: 2026, salesAdjusted: 1350, tagCount: 2 },
  { id: 3, NAME_UPPER: 'CHARLIE', joinYear: 2026, salesAdjusted: 2800, tagCount: 2 },
  { id: 4, NAME_UPPER: 'DAVID', joinYear: 2026, salesAdjusted: null, tagCount: 1 }
]
*/

2. GroupBy & Multi-Metric Aggregations

const transactions = $df.data([
  { region: "East", category: "Hardware", revenue: 450, quantity: 3, timestamp: "2026-02-01" },
  { region: "East", category: "Hardware", revenue: 550, quantity: 5, timestamp: "2026-02-10" },
  { region: "West", category: "Software", revenue: 1200, quantity: 1, timestamp: "2026-01-20" },
]);

const salesSummary = transactions
  .groupBy(["region", "category"])
  .agg([
    $df.col("revenue").sum().alias("total_revenue"),
    $df.col("revenue").mean().round(2).alias("avg_revenue"),
    $df.col("quantity").sum().alias("units_sold"),
    $df.col("timestamp").str.toDatetime().dt.year().alias("year")
  ])
  .sort("total_revenue", { descending: true });

console.log(salesSummary.toDicts());
/* Output:
[
  { region: 'East', category: 'Hardware', total_revenue: 1000, avg_revenue: 500, units_sold: 8, year: 2026 },
  { region: 'West', category: 'Software', total_revenue: 1200, avg_revenue: 1200, units_sold: 1, year: 2026 }
]
*/

3. Relational Joins & Conditional Columns

const orders = $df.data([
  { orderId: 101, customerId: 1, amount: 250 },
  { orderId: 102, customerId: 2, amount: 1400 },
]);

const customers = $df.data([
  { id: 1, firstName: "Alice", lastName: "Smith" },
  { id: 2, firstName: "Bob", lastName: "Jones" },
]);

// Join tables with complete null safety and compute new columns
const enrichedOrders = orders.join(customers, {
  leftOn: "customerId",
  rightOn: "id",
  how: "left"
}).withColumns([
  $df.col("firstName").str.concat(" ").str.concat($df.col("lastName")).alias("fullName"),
  $df.when($df.col("amount").gt(1000)).then("VIP").otherwise("Standard").alias("tier")
]);

console.log(enrichedOrders.toDicts());
/* Output:
[
  { orderId: 101, customerId: 1, amount: 250, id: 1, firstName: 'Alice', lastName: 'Smith', fullName: 'Alice Smith', tier: 'Standard' },
  { orderId: 102, customerId: 2, amount: 1400, id: 2, firstName: 'Bob', lastName: 'Jones', fullName: 'Bob Jones', tier: 'VIP' }
]
*/

4. Rolling Moving Windows & Cumulative Calculations

const sensorData = $df.data([
  { timestamp: "2026-03-01T00:00:00Z", temperature: 20.1, powerUsage: 12 },
  { timestamp: "2026-03-01T01:00:00Z", temperature: 20.9, powerUsage: 14 },
  { timestamp: "2026-03-01T02:00:00Z", temperature: 21.5, powerUsage: 13 },
]);

// Compute moving averages and cumulative metrics without manual loop state
const telemetryWithTrends = sensorData
  .sort("timestamp")
  .withColumns([
    $df.col("temperature").rollingMean(2).alias("temp_rolling_avg"),
    $df.col("powerUsage").cumSum().alias("cumulative_kwh"),
    $df.col("temperature").diff().alias("temp_delta")
  ]);

console.log(telemetryWithTrends.toDicts());
/* Output:
[
  { timestamp: '2026-03-01T00:00:00Z', temperature: 20.1, powerUsage: 12, temp_rolling_avg: 20.1, cumulative_kwh: 12, temp_delta: null },
  { timestamp: '2026-03-01T01:00:00Z', temperature: 20.9, powerUsage: 14, temp_rolling_avg: 20.5, cumulative_kwh: 26, temp_delta: 0.8 },
  { timestamp: '2026-03-01T02:00:00Z', temperature: 21.5, powerUsage: 13, temp_rolling_avg: 21.2, cumulative_kwh: 39, temp_delta: 0.6 }
]
*/

✨ Key Features

📦 Zero Dependencies — 0 external runtime dependencies; completely standalone (<45 KB gzipped).
⚡ Columnar Execution — Flat memory layouts and contiguous loops that eliminate intermediate array allocations.
🔗 Expression-Based API — Compose complex transformations, aggregations, and conditions using fluent $df expressions.
📂 Strict Namespaces:
- .str — Unicode string manipulations, regex extractions, and JSON path lookups.
- .dt — Timezone conversions, business days, microsecond-precision datetimes, and durations.
- .arr — Array/list column operations and element-wise .arr.eval() mapping.
- .struct — Nested object handling and .struct.unnest() column flattening.
🛡️ Defensive & Type-Safe — Automatic type coercion, Kleene three-valued logic for null safety, and strict schema validation.

📌 Compatibility & Runtime Specifications

| Attribute | Specification | | :--- | :--- | | Supported Languages | TypeScript (.ts, .tsx), JavaScript (.js, .jsx), ESM (.mjs), CommonJS (.cjs) | | Target Standard | Standard ES2020+ compatible | | Target Platforms | Node.js (14+), Bun, Deno, Modern Web Browsers (Chrome 80+, Safari 13.1+, Firefox 74+, Edge 80+), Cloudflare Workers, Fastly Compute, AWS Lambda | | UI Frameworks | React, Next.js, Vue, Nuxt, Svelte, SolidJS, Astro, Vite, Webpack, esbuild | | Operational Scope | In-memory columnar DataFrame & vectorized expression engine with zero intermediate allocations | | I/O & Formats | High-throughput RFC-compliant CSV parsing & streaming, JSON / JSONLines, and row/column records | | Runtime Footprint | Pure TypeScript/JavaScript (0 external dependencies, 0 native C++/Rust bindings, 0 WebAssembly overhead) | | Bundle Size | ~44 KB gzipped (~127 KB minified) for the complete engine; ~21 KB gzipped for standalone utils | | Distribution | Dual ESM (dist/index.mjs) & CommonJS (dist/index.js) with comprehensive .d.ts declaration maps | | Package Managers | npm, pnpm, bun, yarn |

💡 Why df-script? (Polars & Pandas for TypeScript)

For the vast majority of web applications, transforming data in a meaningful way still imposes a heavy architectural tax: spinning up a dedicated backend service (commonly Python with Pandas or Polars).

Suddenly, a straightforward transformation requirement introduces container pipelines, deployment overhead, network latency, serialization bottlenecks, UI loading spinners, and ongoing cloud hosting bills, all just to run operations modern client CPUs can crunch in single-digit milliseconds.

When developers attempt to eliminate the backend and transform data directly in JavaScript or TypeScript, they hit another roadblock: hand-rolling bespoke data processing helpers. What begins as a seemingly innocent 10-line utility for grouping, joining, or rolling metrics quickly balloons into hundreds of lines of fragile, bug-ridden boilerplate. Real-world client data immediately exposes an overwhelming mountain of edge cases:

Composite Keys: Hashing multi-column keys without collisions, distinguishing string "null" from literal null and undefined, maintaining deterministic order, and handling sparse indices.
Relational Joins: Preserving outer join semantics without silent truncation, resolving column name collisions with automatic suffixing, and aligning time-series with as-of nearest matches.
Windowing & Rolling Metrics: Managing sliding window boundaries, min-sample thresholds, cumulative accumulators, and lead/lag offsets without unbounded memory growth.
Numeric Gotchas: Eliminating silent precision loss from floating-point drift, NaN contagion, -0 vs +0, integer overflows, and BigInt cross-type operations.
Date & String Edge Cases: Handling timezone offsets across client environments, ISO/RFC parsing variations, regex escaping, and Unicode surrogate pairs.

Attempting to solve these edge cases using naive chains of .map(), .filter(), .flatMap(), and .reduce() compounds the issue: each chained operation performs another full pass over the dataset and floods the V8 garbage collector with millions of short-lived intermediate objects, triggering noticeable UI freezes, frame drops, and memory leaks.

Groupings, relational joins, and rolling windows are only a fraction of the data transformations modern applications need. Hand-rolling solutions for all of them is an unsustainable engineering sinkhole.

df-script changes this equation entirely. It introduces a zero-dependency, column-oriented DataFrame engine constructed specifically to bring native transformation power directly to the browser and edge:

🎯 Target Sweet Spot (1,000 to 500,000+ rows): Purpose-built for operational client datasets, real-time dashboards, local CSV exploration, and IoT telemetry streams directly in-memory or inside a Web Worker.
🚀 Eliminate the Python Backend Tax: Execute complex multi-column groupings, rolling statistics, relational joins, time-series matching, and pivots locally and synchronously right where the user interacts with the data. Zero API round-trips, zero server downtime, and zero backend maintenance.
🛠️ Never Hand-Roll groupBy, join, or Windowing Again: Complex grouping logic, Kleene three-valued null safety, and edge cases are already solved and verified across thousands of unit tests.
🪶 Lightweight Footprint (<45 KB gzipped): Zero external dependencies (~127 KB minified) with no WebAssembly startup latency, no asynchronous compilation steps, and no native C++ bindings.
⚡ Compiled Columnar Performance: Compiles declarative $df.col(...) expression trees down to contiguous, element-wise loops with cached lengths and minimal memory allocations, keeping your UI responsive at 60 FPS.
🛡️ TypeScript-First & Universal: Clean, Polars-inspired domain namespaces (.str, .dt, .arr, .struct) with automatic schema deduction, full compile-time type safety, and standard ES2020+ portability across browsers, Web Workers, Node.js, Bun, and Deno.

📖 Core Concepts

The `$df` Entry Point

df-script uses the $df namespace to instantiate DataFrames, reference columns, construct expressions, and specify data types.

$df.data(dataRowsOrCols, schema?): Instantiates a new DataFrame.
$df.readJson(content, options?): Reads JSON/NDJSON content into a new DataFrame.
$df.readCsv(content, options?): Reads CSV content into a new DataFrame with automatic schema inference.
$df.col(selector): Creates a column reference expression by column name ("a"), multiple names (["a", "b"]), RegExp pattern (/^user_/), or DataType selector ($df.Float64, $df.Numeric).
$df.all(): Selects all columns in the DataFrame.
$df.coalesce(...exprs): Returns the first non-null value among columns or expressions.
$df.concat(items, options?): Concatenates multiple DataFrames along vertical, horizontal, or diagonal axes.
$df.duration(optionsOrString): Constructs a Duration expression from component options ({ days: 1, hours: 12 }) or compound duration strings ("1d 12h 30m").
$df.element(): References the current array element within an .arr.eval(...) expression.
$df.exclude(columns): Matches all columns except the specified ones.
$df.horizontal(...exprs): Combines multiple columns/expressions horizontally per row for row-wise evaluation (.sum(), .arr, .eval()).
$df.implode(column): Aggregates a column's rows or grouped values into a list.
$df.lit(val): Explicitly wraps a raw value into a literal expression.
$df.seqRange(value, options?): Generates a sequence range of values.
$df.struct(fields): Constructs a nested struct object expression from named expressions or sibling columns.
$df.when(predicate).then(value)...otherwise(value): Constructs a conditional CASE WHEN expression chain.
$df.Float64, $df.Int32, $df.Utf8, etc.: Direct access to data types and constructors for schema definitions and type-based column selection.

↑ Back to Top

🛠️ DataFrame API Reference

All methods and getters on DataFrame in alphabetical order:

| Method / Property | Description | | :--- | :--- | | cast(schema, options?) | Casts DataFrame columns to target data types using a schema mapping. | | clone() | Deep-copies the DataFrame, duplicating all underlying columnar memory buffers and schema. | | columns | Getter returning an array of all column names in the DataFrame. | | concat(items, options?) | Concatenates multiple DataFrames along "vertical", "horizontal", or "diagonal" axes. | | drop(...names) | Drops one or more specified columns from the DataFrame. | | dropNulls(subset?) | Drops rows containing null or undefined values across all or specified columns. | | dtypes | Getter returning a dictionary of column names to their registered DataType instances. | | equals(other, options?) | Deep equality comparison with tolerance support for float columns. | | explode(columns) | Unnests list-like columns into multiple rows, replicating other columns per list element. | | fill(target, options?) | Replaces targeted values ("null", "nan", "all", or expressions) across columns with strategies or values. | | fillNan(options?) | Replaces NaN values across numeric columns with scalars, expressions, or statistical strategies. | | fillNull(ruleOrOptions) | Fills null values with strategies ("forward", "backward", "mean", "min", "max", or literal values). | | filter(...predicates) | Filters rows where all predicate expressions evaluate to true. | | groupBy(keys) | Groups data by one or more columns, returning a GroupedData object for aggregations. | | groupByDynamic(indexCol, options) | Dynamic time-series/numeric window grouping over tumbling, sliding, or rolling temporal intervals. | | head(n?) | Returns the first n rows of the DataFrame (defaults to 5). | | height | Getter returning the total number of rows in the DataFrame. | | insertColumn(index, name, values) | Inserts a new column at a specific zero-based index. | | item(row?, col?) | Returns an individual scalar cell value at the specified row and column coordinate. | | iterColumns() | Generator iterating over { name, values } pairs for each column. | | iterRows() | Generator iterating over row record objects { [column]: value }. | | join(other, onOrOptions, how?, suffixes?) | Relational joins ("inner", "left", "right", "outer", "semi", "anti", "cross"). | | joinAsof(other, options) | Inexact time-series / nearest-neighbor joins on sorted key columns. | | joinWhere(other, predicate, options?) | Arbitrary non-equi joins evaluated over candidate cartesian pairs. | | limit(n, options?) | Returns a subset of rows with optional offset and direction ("start" or "end"). | | partitionBy(by, options?) | Splits DataFrame into sub-DataFrames (as array or dictionary) partitioned by column keys or expressions. | | pivot(index, columns, values) | Reshapes tabular data from long to wide format based on unique column keys. | | rename(mapping) | Renames columns using a { oldName: newName } dictionary. | | reverse() | Reverses the row ordering of the DataFrame. | | schema | Getter returning the current column-to-DataType schema specification. | | select(...exprs) | Projects and evaluates column expressions, selectors, regex patterns, or struct unnesting. | | shape | Getter returning the [height, width] dimensions tuple. | | slice(start, end?) | Slices a continuous range of rows by start and end indices. | | sort(options) | Multi-column sorting with descending, nullsLast, and custom comparator functions. | | tail(n?) | Returns the last n rows of the DataFrame (defaults to 5). | | toArray() | Serializes the DataFrame into a two-dimensional row array (any[][]). | | toDict() | Serializes the DataFrame into a column-oriented dictionary { [col]: any[] }. | | toDicts() | Serializes the DataFrame into an array of row objects RowRecord[]. | | transpose(options?) | Swaps rows and columns into a transposed DataFrame. | | unique(columns?) | Returns distinct rows, optionally deduplicating based on a subset of columns. | | unpivot(idVars, valueVars, varName?, valName?) | Melts wide columns into long format name-value pairs. | | unstack(step, options?) | Unstacks multi-level grouped columns. | | width | Getter returning the total number of columns in the DataFrame. | | withColumns(...exprs) | Adds or updates columns in-place using expression evaluations. | | withRowIndex(name?, offset?) | Inserts a zero-based or offset row index column into the DataFrame. | | writeCsv(file?, options?) | Serializes the DataFrame into RFC-4180 CSV string or writes to disk/stream. | | writeJson(file?, options?) | Serializes the DataFrame into JSON or NDJSON string or writes to disk/stream. |

↑ Back to Top

🧮 Column Expressions API Reference

All column expressions inherit from ColumnExpr / ExprBase and support fluent chaining.

1. Core & Math Expressions (Alphabetical)

| Method | Description | | :--- | :--- | | abs | Computes element-wise absolute value. | | add | Element-wise addition with value or expression. | | all | Returns true if all elements evaluate to true (Kleene logic). | | allNull | Returns true if all elements in the expression are null or undefined. | | and | Logical AND with Kleene three-valued null logic. | | any | Returns true if at least one element evaluates to true. | | anyNull | Returns true if any element is null or undefined. | | arccos | Element-wise inverse cosine in radians. | | arccosh | Element-wise inverse hyperbolic cosine. | | arcsin | Element-wise inverse sine in radians. | | arcsinh | Element-wise inverse hyperbolic sine. | | arctan | Element-wise inverse tangent in radians. | | arctan2 | Element-wise four-quadrant inverse tangent with second argument. | | arctanh | Element-wise inverse hyperbolic tangent. | | argMax | Returns the index of the first maximum value. | | argMin | Returns the index of the first minimum value. | | avg | Alias for mean(); computes the arithmetic mean. | | between | Evaluates if elements are within an inclusive or exclusive numeric/date range. | | bitwiseAnd | Bitwise AND (&) operation on integers. | | bitwiseOr | Bitwise OR (|) operation on integers. | | bitwiseXor | Bitwise XOR (^) operation on integers. | | cbrt | Computes element-wise cube root. | | ceil | Rounds numbers upward to the nearest integer. | | clip | Clamps values between a lower and upper boundary. | | copysign | Copies the sign of the magnitude argument to the target numbers. | | corr | Computes Pearson correlation coefficient with another column or expression. | | cos | Element-wise cosine in radians. | | cosh | Element-wise hyperbolic cosine. | | cot | Element-wise cotangent in radians. | | count | Counts non-null elements (or all elements with { includeNulls: true }). | | cov | Computes covariance with another column or expression. | | cumCount | Cumulative count of elements along the column. | | cumMax | Cumulative maximum value along the column. | | cumMin | Cumulative minimum value along the column. | | cumProd | Cumulative product of values along the column. | | cumSum | Cumulative running sum along the column. | | degrees | Converts angles from radians to degrees. | | denseRank | Computes dense rank window order without rank gaps. | | div | Element-wise division. | | dot | Computes dot product vector multiplication with another expression. | | entropy | Computes Shannon entropy (supports optional base and normalization). | | eq | Strict equality (==) with Kleene null propagation. | | eqMissing | Null-safe equality treating null and undefined as equivalent. | | ewmKurt | Computes exponentially weighted moving kurtosis. | | ewmMean | Computes exponentially weighted moving average/mean (supports alpha, span, com, halfLife, and by). | | ewmSkew | Computes exponentially weighted moving skewness. | | ewmStd | Computes exponentially weighted moving standard deviation. | | ewmSum | Computes exponentially weighted moving sum. | | ewmVar | Computes exponentially weighted moving variance. | | exp | Computes natural exponential (e^x). | | expm1 | Computes e^x - 1 with high precision for values near zero. | | fill | Replaces targeted values ("null", "nan", "all", or expressions) with values or propagation strategies. | | fillNan | Replaces NaN values with scalars, expressions, or statistical strategies. | | fillNull | Fills null values with scalar values or forward/backward fill strategies. | | filter | Filters expression elements by a boolean predicate mask. | | first | Returns the first element of the expression or group. | | floor | Rounds numbers downward to the nearest integer. | | floordiv | Integer/floor division. | | ge | Greater than or equal to (>=) comparison. | | gt | Greater than (>) comparison. | | hasNulls | Returns true if column contains at least one null or undefined value. | | hypot | Computes hypotenuse sqrt(a^2 + b^2). | | implode | Aggregates elements into a single list/array column. | | isClose | Evaluates if numeric values are approximately equal within tolerances. | | isDuplicated | Returns boolean mask where true marks duplicated values. | | isFinite | Checks if number is finite (neither NaN, Infinity, nor -Infinity). | | isIn | Checks if value is present in a set or array of values. | | isInfinite | Checks if number is Infinity or -Infinity. | | isNDistinct | Matches the N-th distinct value in order (supports negative index). | | isNan | Checks if numeric value is NaN. | | isNotNan | Checks if numeric value is not NaN. | | isNotNull | Checks if value is neither null nor undefined. | | isNull | Checks if value is null or undefined. | | isUnique | Returns boolean mask where true marks unique (non-duplicated) values. | | kurtosis | Computes fourth standardized moment (peakedness/kurtosis). | | lag | Shifts values backward by offset within partition window. | | last | Returns the last element of the expression or group. | | le | Less than or equal to (<=) comparison. | | lead | Shifts values forward by offset within partition window. | | log | Computes logarithm with optional base (defaults to natural log). | | log1p | Computes natural log of (1 + x) for values near zero. | | lt | Less than (<) comparison. | | max | Returns the maximum non-null value. | | maxBy | Returns the value of this expression where another expression achieves maximum. | | mean | Computes the arithmetic average of non-null numbers. | | median | Computes the 50th percentile median value. | | min | Returns the minimum non-null value. | | minBy | Returns the value of this expression where another expression achieves minimum. | | mod | Modulo division remainder. | | mode | Computes the most frequently occurring value(s). | | mul | Element-wise multiplication. | | nUnique | Returns the count of distinct non-null values. | | nanMax | Returns maximum treating NaN values according to IEEE 754 rules. | | nanMin | Returns minimum treating NaN values according to IEEE 754 rules. | | ne | Strict inequality (!=) with Kleene null propagation. | | neMissing | Null-safe inequality. | | negate | Negates numeric values (-x). | | not | Logical NOT operator. | | notIn | Checks if value is not present in a set or array. | | nullCount | Returns the total count of null and undefined values. | | or | Logical OR with Kleene three-valued null logic. | | over | Evaluates the expression over an analytical partition window (grouping keys). | | pow | Computes exponentiation (x^y). | | product | Computes product of all non-null values. | | quantile | Computes quantile value at probability p (0 <= p <= 1). | | radians | Converts angles from degrees to radians. | | rand | Generates uniform pseudo-random values between 0 and 1. | | rank | Computes competition rank order (1, 2, 2, 4). | | reverse | Reverses the element order of the column. | | rolling | Evaluates custom function or expression over a sliding window. | | rollingMax | Computes rolling maximum over a sliding window. | | rollingMean | Computes rolling moving average over a sliding window. | | rollingMedian | Computes rolling median over a sliding window. | | rollingMin | Computes rolling minimum over a sliding window. | | rollingQuantile | Computes rolling quantile over a sliding window. | | rollingRank | Computes rolling rank over a sliding window. | | rollingStd | Computes rolling standard deviation over a sliding window. | | rollingSum | Computes rolling sum over a sliding window. | | round | Rounds numbers to specified decimal places. | | roundSigFigs | Rounds numbers to a specified number of significant figures. | | rowNumber | Computes sequential 1-based row index within partition window. | | shift | Shifts column values by a signed offset (filling empty cells with null). | | sign | Extracts sign of number (-1, 0, 1). | | sin | Element-wise sine in radians. | | sinh | Element-wise hyperbolic sine. | | skew | Computes third standardized moment (skewness asymmetry). | | spearmanCorr | Computes Spearman rank correlation coefficient. | | sqrt | Computes square root. | | std | Computes sample standard deviation. | | sub | Element-wise subtraction. | | sum | Computes sum of all non-null numeric values. | | tan | Element-wise tangent in radians. | | tanh | Element-wise hyperbolic tangent. | | trunc | Truncates fractional parts toward zero. | | variance | Computes sample variance. | | wAvg | Computes weighted average using weight expression. | | xor | Logical exclusive OR (XOR). |

↑ Back to Top

📂 Namespaces API Reference

Specific domain transforms are grouped under dedicated, clean namespaces with full IDE autocomplete:

🔤 String Namespace (`.str`)

📅 Temporal Namespace (`.dt`)

📊 Array/List Namespace (`.arr`)

🗃️ Struct/Object Namespace (`.struct`)

↑ Back to Top

🛡️ Typing and Schema Registry

import { $df } from "df-script";

const schema = {
  id: $df.DataType.Int32,
  price: $df.DataType.Decimal(10, 2),
  active: $df.DataType.Boolean,
  createdAt: $df.DataType.Datetime
};

const df = $df.data(rawData, schema);

Supported Data Types

Integers: Int8, Int16, Int32, Int64, UInt8, UInt16, UInt32, UInt64
Floats & Decimals: Float32, Float64, Decimal(precision?, scale?)
General: Boolean, Utf8 (Strings), Binary, Null, Object
Temporal: Date, Datetime, Time, Duration
Nested Structures: List (Arrays), Struct (Objects)

↑ Back to Top

🧑‍💻 Contributing & Development

# Run test suite
npm test

# Build production bundles
npm run build

📄 License

df-script is open-source software licensed under the MIT License.

Published

Vulnerabilities

Links

Maintainers

Keywords

Readme

🚀 df-script: High-Performance TypeScript DataFrame Library

📦 Installation

🚀 Quick Start

1. Basic DataFrame & Expression Pipeline

2. GroupBy & Multi-Metric Aggregations

3. Relational Joins & Conditional Columns

4. Rolling Moving Windows & Cumulative Calculations

✨ Key Features

📌 Compatibility & Runtime Specifications

💡 Why df-script? (Polars & Pandas for TypeScript)

📖 Core Concepts

The $df Entry Point

🛠️ DataFrame API Reference

🧮 Column Expressions API Reference

1. Core & Math Expressions (Alphabetical)

📂 Namespaces API Reference

🔤 String Namespace (.str)

📅 Temporal Namespace (.dt)

📊 Array/List Namespace (.arr)

🗃️ Struct/Object Namespace (.struct)

🛡️ Typing and Schema Registry

Supported Data Types

🧑‍💻 Contributing & Development

📄 License

The `$df` Entry Point

🔤 String Namespace (`.str`)

📅 Temporal Namespace (`.dt`)

📊 Array/List Namespace (`.arr`)

🗃️ Struct/Object Namespace (`.struct`)