qntjs-lib 1.0.0 → 1.0.1

package/README.md

@@ -1,17 +1,17 @@
 # qntjs-lib
 
-A compact JavaScript/TypeScript library of numerical helpers and technical-analysis indicators.
+High-performance JavaScript and TypeScript library of technical‑analysis indicators, array utilities, and numerical helpers.
 
-This repository provides a comprehensive set of TA indicators (EMA, TEMA, T3, MFI, KAMA, etc.), vectorized math functions, array utilities, and statistical helpers.
+This repository implements a wide set of TA indicators (EMA, TEMA, T3, MFI, KAMA, etc.), vectorized math functions, and statistical helpers. Implementations are optimized for two common usage patterns:
 
-For best performance, functions operate and return typed arrays like Float64Array.
-For convenience, the same functions are also exposed with number[] return types.
+- NaN‑aware workflows (default): functions are NaN‑aware and will skip NaN values where appropriate.
+- Dense fast‑path: when you know inputs contain no NaNs you can opt into a dense, faster implementation by passing `skipna=false` to supported functions.
 
-Most functions skip NaN by default in their calculations. 
-Some functions accept boolean skipna to skip or not NaN.
-When skipna=false is passed to these functions, a fast dense path is executed. 
+By default the main (typed) build returns typed arrays (e.g. `Float64Array`) for better numeric performance and predictable memory layout. A companion "untyped" build exposes the same API but returns plain `number[]` values for easier interoperability with plain JavaScript code.
 
-**Quick Start**
+The library has no runtime dependencies and is zero-dependency by design. It can be used in browser web applications or in Node.js environments that support ESM imports.
+
+## Quick Start
 
 Install:
 
@@ -19,23 +19,50 @@ Install:
 npm install qntjs-lib
 ```
 
-Basic usage:
+Basic usage (default — typed output):
+
+```js
+import { ta } from 'qntjs-lib';
+
+const prices = [1,2,3,4,5,6,7];
+// returns Float64Array by default (typed numeric output)
+const out = ta.ema(prices, 3);
+```
+
+Basic usage (untyped — plain arrays):
 
 ```js
-import { ta, arr } from 'qntjs-lib';
+import { ta } from 'qntjs-lib/untyped';
 
 const prices = [1,2,3,4,5,6,7];
+// returns plain number[] (easier to inspect/serialize)
 const out = ta.ema(prices, 3);
-console.log(out);
 ```
 
-API highlights
+When to use each:
+- Use the default import (`qntjs-lib`) when you want outputs as `Float64Array` for numeric performance and predictable memory layout.
+- Use `qntjs-lib/untyped` when you prefer plain `number[]` outputs for easier inspection or serialization.
+
+## skipna and dense fast-path
+
+Many indicators accept an optional `skipna` boolean (default `true`). Example:
+
+```js
+// NaN-aware (default)
+ta.sma(pricesWithGaps, 5);
+
+// Dense fast-path (assume no NaNs)
+ta.sma(densePrices, 5, false);
+```
+
+## API highlights
 
-- `ta.*` — technical indicators (e.g. `ta.ema`, `ta.t3`, `ta.mfi`, `ta.kama`)
-- `arr.*` — array utilities
-- `math.*` — math helpers
+ - `arr.*` — array utilities: `isna`, `notna`, `fillna`, `ffill`, `bfill`, `replace`, `dropna`, `allna`, `equals`, `countna`, `havena`, `lag`
+ - `math.*` — math helpers: `add`, `sub`, `avg`, `mul`, `div`, `scale`, `abs`, `sign`, `round`, `floor`, `ceil`, `eq`, `neq`, `gt`, `gte`, `lt`, `lte`, `and`, `or`, `not`, `clamp`, `sum`, `prod`, `min`, `max`, `argmin`, `argmax`, `cumsum`, `cumprod`, `cummax`, `cummin`, `rollsum`, `rollmin`, `rollmax`, `rollminmax`, `rollprod`, `rollargmin`, `rollargmax`, `diff`, `randuniform`, `randnormal`, `dot`, `norm`, `ols`, `olsMulti`
+ - `stats.*` — statistical helpers: `mean`, `hmean`, `gmean`, `mad`, `skew`, `kurtosis`, `median`, `quantile`, `percentiles`, `var`, `covar`, `stdev`, `corr`, `zscore`, `norminmax`, `winsorize`, `sample`, `shuffle`, `bootstrap`
+- `ta.*` — technical indicators: `dema`, `ema`, `hma`, `kama`, `sma`, `wma`, `vwma`, `trima`, `t3`, `tema`, `rma`, `ao`, `apo`, `aroon`, `change`, `cmo`, `kst`, `macd`, `mom`, `ppo`, `roc`, `rsi`, `stoch`, `stochrsi`, `ultosc`, `wpr`, `supertrend`, `adx`, `adxr`, `dx`, `cci`, `di`, `dpo`, `ichimoku`, `psar`, `atr`, `tr`, `natr`, `bb`, `bbw`, `donchian`, `keltner`, `adosc`, `obv`, `pnvi`, `wad`, `ad`, `mfi`, `cross`, `crossover`, `crossunder`, `rising`, `falling`
 
-Tests & development
+## Tests & development
 
 Run tests:
 
@@ -43,18 +70,18 @@ Run tests:
 npm test
 ```
 
-Run build (produce distributable artifacts):
+Build (produce distributable artifacts):
 
 ```bash
 npm run build
 ```
 
-Contributing
+## Contributing
 
 - Open an issue to discuss larger changes.
 - Run tests and keep coverage high for changes to core algorithms.
 - Follow the existing style and types in `src/`.
 
-License
+## License
 
 See `LICENSE` in the repo root.

package/package.json

@@ -1,6 +1,6 @@
 {
   "name": "qntjs-lib",
-  "version": "1.0.0",
+  "version": "1.0.1",
   "type": "module",
   "main": "./dist/bundle/index.js",
   "types": "./dist/index.d.ts",
@@ -36,7 +36,21 @@
     "type": "git",
     "url": "git+https://github.com/amahouachi/qntjs-lib.git"
   },
-  "keywords": ["technical-analysis","indicators","ta","trading","finance","sma","ema","tema","kama","mfi","typescript","javascript","array-utils"],
+  "keywords": [
+    "technical-analysis",
+    "indicators",
+    "ta",
+    "trading",
+    "finance",
+    "sma",
+    "ema",
+    "tema",
+    "kama",
+    "mfi",
+    "typescript",
+    "javascript",
+    "array-utils"
+  ],
   "author": "Ahmed MAHOUACHI (https://github.com/amahouachi)",
   "license": "MIT",
   "bugs": {