a2ff35f5f92e4b95dc024ff0bd8cbc5e96f53c32
244 Commits
| Author | SHA1 | Message | Date | |
|---|---|---|---|---|
 kingchenc
|
a2ff35f5f9 |
ci(sync-about): auto-sync repo homepage + org profile from the indicator count (#84)
Extend the count-sync workflow to drive three more surfaces from the same
single count, so the org page never drifts again:
- Repo About **homepage** URL — enforced unconditionally via
`gh repo edit --homepage` in the existing About step (fixes the stale
kingchenc URL and self-heals). Uses the same Administration-write
permission as --description, so NO extra token scope.
- Org **profile README** count (wickra-lib/.github, profile/README.md) —
clone + sed + bot commit/push, same pattern as the Sync Wiki step.
- Org **description** field ('… N indicators, install-free.') — gh api PATCH.
The two org steps need ABOUT_SYNC_TOKEN scope the main-repo syncs do not
(write on wickra-lib/.github; admin:org for the description PATCH). Both are
written to soft-skip with a ::warning:: and continue-on-error, so the run
stays green until the scope is granted — then they sync with no further
change. Homepage swap-point to a docs domain is a one-line change.
Refs findings P9 / P10.
|
||
|
kingchenc
|
01aeb965d1 |
release: bump 0.3.0 -> 0.3.1 (#80)
* chore: remove ROADMAP.md from the public repo ROADMAP is kept as a local-only draft (ghost-ignored via .git/info/exclude); it is not part of the published package surface. * release: bump 0.3.0 -> 0.3.1 CI-only patch: fixes the release.yml CycloneDX SBOM step (cargo-cyclonedx has no -p flag, see #79) that skipped the GitHub Release attach-assets job on 0.3.0. No library changes — republishes the same code with a working release pipeline. - Cargo.toml (workspace.package + wickra-core dep) + Cargo.lock - bindings/python/pyproject.toml - bindings/node/package.json (version + 6 optionalDependencies) + package-lock.json - bindings/node/npm/*/package.json (6 platform subpackages) - CHANGELOG: finalize [0.3.0] (was still under [Unreleased]), add [0.3.1] * chore: track examples/node/package-lock.json Since the global package-lock ignore rule was dropped (#68) this file was left untracked. Commit it for reproducible example installs, consistent with bindings/node (findings P4.1).v0.3.1 |
||
|
kingchenc
|
f1fed6cdd5 |
ci(release): fix CycloneDX SBOM generation (cargo-cyclonedx has no -p flag) (#79)
cargo-cyclonedx 0.5.9 walks the whole workspace in a single pass and writes a <package>.cdx.json next to each member's Cargo.toml; it has no -p/--package selector. The previous three 'cargo cyclonedx ... -p <crate>' invocations aborted with 'error: unexpected argument -p found', failing the cargo-publish job after the crates were already published and, in turn, skipping the github-release attach-assets job (it needs all four publish jobs). v0.3.0 published to every registry but got no GitHub Release page or SBOM assets as a result. Replace the three invalid calls with a single workspace pass and copy the three crates.io crate SBOMs into the upload dir. |
||
|
kingchenc
|
70e9cbb397 |
release: bump 0.2.7 -> 0.3.0 (supersedes PR #61) (#69)
Minor bump (not patch) because the [Unreleased] section since 0.2.7 has accumulated a sweep of additive changes that justify a new minor: - Family 9-16 indicator catalogue expansion (Bands & Channels, Trailing Stops, Volume, Statistics, Ehlers/Cycle DSP, Pivots, DeMark, Ichimoku, Candlestick Patterns, Market Profile, Risk/Performance) — roughly 100+ new indicators since 0.2.7 across all four bindings. - New `wickra_core::FAMILIES` const + family-taxonomy guard tests. - GitHub org migration (kingchenc -> wickra-lib) and new maintainer email (wickra.lib@gmail.com). - New `repo-metadata.toml` + `sync-metadata.yml` audit workflow. - WASM CI tests now run on every PR (existing tests had been manually-only). - CycloneDX SBOMs + npm provenance attestations attached to releases. - Three end-to-end strategy examples. - Governance polish: ARCHITECTURE / ROADMAP / CITATION / FUNDING / .editorconfig. - Curated benchmark suite (~33 representative indicators). - Three cold-path coverage fixes (mama, rsi, sine_wave). - bindings/node/package-lock.json now committed. Workspace + bindings (Rust crate, Python wheel, Node main + 6 platform sub-packages, WASM) all step to 0.3.0. CHANGELOG opens the [0.3.0] section dated 2026-05-28 with the full Changed / Added inventory. Compare-URL block adds the v0.2.7...v0.3.0 line under [Unreleased] and points [Unreleased] at v0.3.0...HEAD using the new wickra-lib org. **Supersedes PR #61** (0.2.7 -> 0.2.8 patch bump). Close #61 when this one merges. Merge ordering remains: #59 (org migration) + #60 (family-api) + the polish PRs first, rebase this PR on top of the new main, then merge. Tag-push `v0.3.0` is a SEPARATE, manual step after merge — it triggers release.yml's irreversible publish to crates.io / PyPI / npm.v0.3.0 |
||
|
dependabot[bot]
|
37e5e19b57 |
deps(actions): bump taiki-e/install-action from 2.79.5 to 2.79.15 (#76)
Bumps [taiki-e/install-action](https://github.com/taiki-e/install-action) from 2.79.5 to 2.79.15. - [Release notes](https://github.com/taiki-e/install-action/releases) - [Changelog](https://github.com/taiki-e/install-action/blob/main/CHANGELOG.md) - [Commits](https://github.com/taiki-e/install-action/compare/6c1f7cf125e42770ff087ea443901b487cc5471a...0fd46367812ee04360509b4169d9f659d6892bb2) --- updated-dependencies: - dependency-name: taiki-e/install-action dependency-version: 2.79.15 dependency-type: direct:production update-type: version-update:semver-patch ... Signed-off-by: dependabot[bot] <support@github.com> Co-authored-by: dependabot[bot] <49699333+dependabot[bot]@users.noreply.github.com> |
||
|
dependabot[bot]
|
c189075491 |
deps(actions): bump EmbarkStudios/cargo-deny-action (#74)
Bumps [EmbarkStudios/cargo-deny-action](https://github.com/embarkstudios/cargo-deny-action) from 2.0.19 to 2.0.20. - [Release notes](https://github.com/embarkstudios/cargo-deny-action/releases) - [Commits](https://github.com/embarkstudios/cargo-deny-action/compare/a531616d8ce3b9177443e48a1159bc945a099823...bb137d7af7e4fb67e5f82a49c4fce4fad40782fe) --- updated-dependencies: - dependency-name: EmbarkStudios/cargo-deny-action dependency-version: 2.0.20 dependency-type: direct:production update-type: version-update:semver-patch ... Signed-off-by: dependabot[bot] <support@github.com> Co-authored-by: dependabot[bot] <49699333+dependabot[bot]@users.noreply.github.com> |
||
|
dependabot[bot]
|
74dd82a867 |
deps(actions): bump actions/checkout from 4 to 6 (#75)
Bumps [actions/checkout](https://github.com/actions/checkout) from 4 to 6. - [Release notes](https://github.com/actions/checkout/releases) - [Commits](https://github.com/actions/checkout/compare/v4...v6) --- updated-dependencies: - dependency-name: actions/checkout dependency-version: '6' dependency-type: direct:production update-type: version-update:semver-major ... Signed-off-by: dependabot[bot] <support@github.com> Co-authored-by: dependabot[bot] <49699333+dependabot[bot]@users.noreply.github.com> |
||
|
kingchenc
|
b654db312e |
docs: fix unresolved and private intra-doc links (#78)
A full `cargo doc --workspace` (and docs.rs, which builds with `-D rustdoc::all`) emitted five broken intra-doc links. docs.rs treats these as hard errors, so any 0.2.x doc build was at risk of aborting. - mama.rs: `[`Fama`]` -> `[`crate::Fama`]` (Fama lives in another module) - standard_error.rs: `[`crate::Bollinger`]` -> `[`crate::BollingerBands`]` (the public type is `BollingerBands`, not `Bollinger`) - aggregator.rs: drop the link to the private `OpenBar::into_candle`, keep it as plain code text - resample.rs: drop the link to the private `RolledBar::into_candle` - csv.rs: `CandleReader::with_timestamp_parser` never existed; reword to state plainly that ISO/RFC timestamps must be converted to integers Verified with `RUSTDOCFLAGS="-D rustdoc::broken_intra_doc_links \ -D rustdoc::private_intra_doc_links" cargo doc --workspace --no-deps`: clean, zero warnings. |
||
|
kingchenc
|
88f119109d |
chore(node): commit bindings/node/package-lock.json for reproducible installs (#68)
Until now `package-lock.json` was globally ignored. Two practical consequences for the Node binding: - A fresh `git clone && cd bindings/node && npm install` resolved `@napi-rs/cli` and any transitive deps to whatever the npm registry currently considered the latest matching the package.json semver ranges. Contributors could get different dep graphs on different days. - No protection against transitive-dep tampering at install time (lockfile records resolved versions + integrity hashes, npm verifies on subsequent installs). Drop the global `package-lock.json` ignore and commit the freshly generated `bindings/node/package-lock.json` (140 lines, only a couple of direct deps because the binding is small). The `.gitignore` comment notes that we still don't expect lockfiles at the workspace root. No CI workflow changes — the existing `npm install` in the Node-test job will now consume the committed lockfile, which is exactly what we want. |
||
|
kingchenc
|
2945b47e1a |
feat(release): add CycloneDX SBOMs and npm provenance attestations (#66)
Two modern supply-chain-trust additions to the release pipeline, neither of which changes what gets published — only adds verifiable signals attached to existing releases. 1. **CycloneDX SBOMs.** `cargo-cyclonedx` is installed in the cargo-publish job after the .crate files are built, and runs once per published crate (`wickra-core`, `wickra-data`, `wickra`). The resulting `*.cdx.json` files are uploaded as the `sboms` artifact, then attached to the GitHub Release alongside the existing wheels, tarballs, .node binaries and .crate files. Future security advisories can answer "is my version of crate X transitive in wickra Y.Z?" by reading the SBOM directly instead of resolving the lockfile. 2. **npm `--provenance` flag** on every npm publish call: - main `wickra` package (node-publish, first + retry) - per-platform `wickra-<triple>` subpackages (node-publish loop) - `wickra-wasm` (wasm-publish) Provenance attestations are generated server-side by npm from the GitHub Actions OIDC token. The publishing jobs gain `permissions: id-token: write` so the runner can exchange that token. The npm page for each published version will then carry the "Verified provenance" badge, which proves the tarball was built by *this* workflow run and not by an arbitrary local laptop with the NPM_TOKEN. Skipped deliberately (to keep this PR focused, possibly follow-ups): - Sigstore cosign signing of artefacts (different audit story; can be layered on after npm-provenance lands). - SLSA build-provenance attestations via `actions/attest-build-provenance` (would target every artefact uniformly; the npm-provenance flag is the more pragmatic first cut). YAML structure validated (8 jobs intact). No production code touched. This PR conflicts with PR #59 only in line-by-line URL substitutions on release.yml — rebase after #59 should be clean. |
||
|
kingchenc
|
c212f91256 |
docs(examples): add 3 end-to-end strategy examples (Rust + Python) (#65)
Wires real indicators into complete signal -> fill -> PnL -> equity loops over the checked-in BTCUSDT datasets, with per-trade Sharpe and max-drawdown reported on stdout. Closes the gap where existing examples showed only the mechanics of calling `update`/`batch` but not how Wickra plugs into a trading-system shape. Three strategies, each in Rust + Python (six files total): - strategy_rsi_mean_reversion — RSI(14) thresholds (30/70) on 1h BTCUSDT. Binary position, 0.1% per-trade fee. - strategy_macd_adx — MACD crossover entries gated by ADX(14) > 20 on 1h BTCUSDT. Trend-follower demo of multi-indicator gating. - strategy_bollinger_squeeze — Bollinger-bandwidth 180-day-low squeeze + upper-band breakout entry, ATR(14) * 2 stop. On 1d BTCUSDT for interpretable lookback. Each file is self-contained — print_summary is inlined per script so the example stays a single-file read. Every script prints a NOT-financial-advice notice next to its results. examples/README.md updated to list the new bins/scripts. |
||
|
kingchenc
|
6c2ddf319f |
feat(wasm): wire wasm-bindgen-test into CI and expand coverage (#64)
* feat(wasm): wire wasm-bindgen-test into CI and expand binding coverage
The WASM binding already had 8 wasm_bindgen_test cases checked in but
they ran only when someone manually invoked `wasm-pack test` locally —
CI never executed them, so a breakage between the Rust core API and the
JS surface would only surface in user reports.
Two changes:
1. **New CI job `wasm-test`** in `.github/workflows/ci.yml` runs
`wasm-pack test --node bindings/wasm` on every push and pull-request.
Uses Node-runtime mode (no headless browser needed), pins
`taiki-e/install-action` for `wasm-pack`, installs the
`wasm32-unknown-unknown` target.
2. **10 new `#[wasm_bindgen_test]` cases** in `bindings/wasm/src/lib.rs`
covering families that the existing 8 tests did not touch:
- Family 4 — Macd multi-output batch shape (3-component packing)
- Family 5 — Bollinger {upper, middle, lower} ordering invariant
- Family 10 — FisherTransform streaming roundtrip (recursive DSP)
- Family 16 — SharpeRatio reset semantics, MaxDrawdown monotone-
uptrend invariant, ValueAtRisk constructor validation
- Cross-family — reset returns indicator to warmup (3 shapes),
warmup_period parity with wickra-core, NaN input rejection
does not advance warmup counter
No production code changed; additive tests only. Total wasm test count
goes from 8 to 18.
* fix(wasm-tests): WasmAtr's hand-coded wrapper has no is_ready accessor
The new `reset_returns_indicator_to_warmup` test in this branch called
`atr.is_ready()` but `WasmAtr` is hand-coded (not generated by the
`wasm_scalar_indicator!` macro) and does not expose `is_ready` on the
JS surface — that pattern is reserved for macro-generated scalar
wrappers.
Replace the second leg of the test with `WasmEma` so the test exercises
two macro-generated scalars whose `is_ready`/`reset` semantics are
guaranteed by the same code path. The candle-input lifecycle is
already covered by `candle_input_streaming_matches_batch_and_lifecycle`.
Workspace clippy locally green:
- cargo clippy -p wickra-wasm --all-targets -- -D warnings
- cargo check -p wickra-wasm --target wasm32-unknown-unknown --tests
* fix(wasm-tests): Bollinger batch layout is 4 floats per bar, not 3
The new `bollinger_batch_orders_upper_mid_lower` test indexed the batch
output as `[u0, m0, l0, u1, m1, l1, ...]` (3 floats per bar) but the
actual WasmBb::batch layout is `[u0, m0, l0, sd0, u1, m1, l1, sd1, ...]`
— four floats per bar including stddev. The assertion read an upper
band against a middle from the previous bar, which can violate the
expected upper >= middle ordering and made the test fail intermittently.
Switch the stride from `* 3` to `* 4` so we read upper / middle / lower
from the same bar. Also add an `is_finite()` precondition so the test
fails with a clear "warmup positions unexpectedly NaN" message rather
than a confusing NaN-comparison if the dataset ever shrinks.
* fix(wasm-tests): drop redundant wasm-test job; existing WASM build covers it
The CI workflow already had a `WASM build` job that runs
`wasm-pack test --node bindings/wasm` after building. The
`wasm-test` job I added in
|
||
|
kingchenc
|
9db1ff8023 |
chore(bench): curate benchmarks to ~30 representative indicators (#63)
Previously every push of `cargo bench -p wickra` ran 114 indicators at three workload sizes each (1k / 10k / 50k candles), which inflated bench runtime past ten minutes for diminishing signal — most family members are linear scalings of the same hot loop, so a regression in any one of them shows up identically in the cheapest member. This commit replaces the exhaustive list with a curated selection: the cheapest baseline and the most expensive representative from each of the sixteen families, totalling ~33 indicators across scalar, candle, and multi-output APIs. If you need to profile a specific indicator that is not in the curated set, add it temporarily and run `cargo bench -- <name>` to target just that bench; it does not need to be committed. Fixed in passing: - `Cci` is `Indicator<Input = Candle>`, not f64; corrected the bench selector to `bench_candle_input`. - `Psar::new` takes (af_start, af_step, af_max) — supplied all three. - `TdSequential` uses `::classic` for the textbook (4, 9, 2, 13) parameters; the old call was missing arguments. |
||
|
kingchenc
|
fb6eae7fe2 |
docs: add ARCHITECTURE, ROADMAP, CITATION, FUNDING, editorconfig (#62)
Five governance + onboarding files that collectively close the "no high-level documentation outside of README" gap. - ARCHITECTURE.md: workspace layout, Indicator trait contract, per-indicator file conventions, numerical-stability notes, cross-crate flow diagram, navigation cheatsheet, deliberate non-goals, performance characteristics, stability commitments. - ROADMAP.md: north star, 0.3 / 0.4 / 0.5 release windows, indicator-wishlist intake rules, explicit non-goals, versioning policy. - CITATION.cff: GitHub-renders as "Cite this repository" button; enables academic adoption. - .github/FUNDING.yml: surfaces Sponsor button on the repo page. - .editorconfig: normalises indent/EOL across IDEs (Rust/Python 4 spaces, JS/TS/JSON/YAML/MD 2 spaces, Makefile tabs). Touches no Rust source, no CI workflows, no behaviour. Additive only. URLs in ROADMAP/CITATION already reference the post-transfer `wickra-lib/wickra` org so the files stay valid once the org migration in PR #59 lands; merge this PR only after #59 to keep main consistent. |
||
|
kingchenc
|
0edb9f4857 |
fix(bindings): clear clippy pedantic lints and lint bindings in CI (#77)
* fix(bindings): clear clippy pedantic lints and lint bindings in CI Resolve the pedantic lints that only surfaced under a full `cargo clippy --workspace` (the CI clippy job covered only the core crates, so the Python/Node bindings drifted): - manual_midpoint: `(a + b) / 2.0` -> `f64::midpoint(a, b)` (node + python) - new_without_default: add `Default` impls for the six no-arg Node nodes - type_complexity: factor the pivot/Ichimoku return tuples into `PivotLevels` / `WoodieLevels` / `IchimokuLines` aliases (python) - many_single_char_names: allow at crate level — OHLCV batch helpers bind the conventional o/h/l/c/v column names Add a dedicated `clippy-bindings` CI job (ubuntu-only, with Python + Node toolchains) so future binding lints fail CI instead of slipping through. * ci: lint Python and Node bindings in a dedicated clippy job The main `rust` job's clippy step only covers wickra-core/wickra/ wickra-data/wickra-wasm, so pedantic lints in the PyO3/napi bindings slipped through. Add an ubuntu-only `clippy-bindings` job that provisions Python + Node (needed by the build scripts) and runs `cargo clippy -p wickra-node -p wickra-python --all-targets -- -D warnings`. |
||
|
kingchenc
|
ea684e0d48 |
feat(family-api): add FAMILIES const and family-taxonomy tests (#60)
* feat(family-api): add FAMILIES const and family-taxonomy tests Introduce `wickra_core::FAMILIES`, a `&'static [(&str, &[&str])]` mapping every built-in indicator to one of 16 families (Moving Averages, Momentum Oscillators, Trend & Directional, Price Oscillators, Volatility & Bands, Bands & Channels, Trailing Stops, Volume, Price Statistics, Ehlers / Cycle (DSP), Pivots & S/R, DeMark, Ichimoku & Charts, Candlestick Patterns, Market Profile, Risk / Performance). Two compile-time-anchored guards make sure the const stays trustworthy: - `no_duplicates_across_families`: no indicator is listed in two families - `total_count_matches_expected`: hard-coded total bumps in lockstep with the indicator catalogue (214 today), so any new indicator added without being filed under a family trips the test. Also corrects the module doc comment which falsely claimed indicators were grouped by category internally; the `mod` block is alphabetical, and the canonical taxonomy now lives in `FAMILIES`. * chore: sync indicator count to 214 --------- Co-authored-by: wickra-bot <wickra-bot@users.noreply.github.com> |
||
|
kingchenc
|
62ab84c472 |
chore(migration): switch org to wickra-lib and maintainer email to wickra.lib@gmail.com (#59)
Introduce repo-metadata.toml as single source of truth for repo identity (org slug, maintainer email, canonical URLs) and add sync-metadata.yml workflow with a Python audit script that fails CI if any tracked file drifts back to pre-migration values. Bulk-replace across 24 tracked files: - kingchenc/wickra -> wickra-lib/wickra (URL segment) - kingchencp@gmail.com -> wickra.lib@gmail.com (maintainer email) - @kingchenc -> @wickra-lib (CODEOWNERS mention only) Person-name credits are preserved: LICENSE copyright holder, Cargo.toml authors handle, and CHANGELOG historical @kingchenc reference all remain unchanged. Crate / PyPI / npm package names also untouched. Merge this PR only after the kingchenc/wickra -> wickra-lib/wickra org transfer has happened on the GitHub side, otherwise all badges and repository links 404 until the transfer is performed. |
||
|
kingchenc
|
0f2ff9c3c7 |
chore(sync-about): count public indicator types, not module files (#70)
* chore(sync-about): count public indicator types, not module files
The sync-about workflow counted `mod xxx;` lines in
crates/wickra-core/src/indicators/mod.rs to derive the indicator count
that gets propagated to README, the GitHub About description, and the
wiki. That under-reported by one because `vwap.rs` exports two public
types — `Vwap` (cumulative) and `RollingVwap` (finite window) — from
the same module. The bindings reach both, so users see 214 indicators
even though there are only 213 source files.
Fix the count by parsing the canonical `pub use indicators::{ ... }`
block in lib.rs, dropping the `FAMILIES` constant and any `*Output`
companion structs, and counting the remaining public types. Pure-shell
implementation so the workflow doesn't grow a python dependency.
Sync README to the corrected count (213 -> 214) in the same commit so
the PR validates cleanly through the workflow's PR-flow.
* docs(bindings): sync per-binding READMEs and docs/ pointer to 214 / 16
The bindings/{node,python,wasm}/README.md files (which ship to
npm / PyPI / npm again) and docs/README.md (the pointer to the
wiki) all carried the stale '71 streaming-first indicators across
eight families' header from before the family expansion. Pull the
canonical 16-family table out of the main README into all three
binding READMEs and update docs/README.md to list every family by
name, so a user landing on PyPI / npm sees the current catalogue
shape.
|
||
|
kingchenc
|
e85334a2e9 |
test(cold-paths): cover 3 lines in mama/rsi/sine_wave (#58)
* test(cold-paths): cover phase fallback in mama/sine_wave and rsi naive helper saturation * fix(rsi): drop redundant closure in test helper |
||
|
kingchenc
|
4e3c41ea80 |
feat(family-15): add 17 risk/performance metrics (#54)
* feat(family-15): add 17 risk/performance metrics Implements Family 15 pragmatically as standard `Indicator`s instead of a separate `wickra-metrics` crate. Input is scalar `f64` per bar — period return, equity sample, or per-trade P&L depending on the metric. Scalar `Indicator<f64>` (14): - SharpeRatio(period, risk_free) - SortinoRatio(period, mar) - CalmarRatio(period) - OmegaRatio(period, threshold) - MaxDrawdown(period) — rolling, peak-to-trough - AverageDrawdown(period) - DrawdownDuration — cumulative, bars under water (u32 output) - PainIndex(period) - ValueAtRisk(period, confidence) - ConditionalValueAtRisk(period, confidence) - ProfitFactor(period) - GainLossRatio(period) - RecoveryFactor — cumulative, net return / max drawdown - KellyCriterion(period) Two-series `Indicator<(f64, f64)>` for (asset, benchmark) returns (3): - TreynorRatio(period, risk_free) - InformationRatio(period) - Alpha(period, risk_free) — Jensen / CAPM Touchpoints: - 17 new files under `crates/wickra-core/src/indicators/`. - `mod.rs` + `lib.rs` re-exports. - Python bindings (`bindings/python/src/lib.rs`, `__init__.py`). - Node bindings (`bindings/node/src/lib.rs`, `index.js`). - WASM bindings (`bindings/wasm/src/lib.rs`). - Fuzz: scalar metrics appended to `indicator_update.rs`; new `indicator_update_pair.rs` fuzz target for `(f64, f64)` indicators. - Python tests: SCALAR + new PAIR parameter lists in `test_new_indicators.py`, reference-value cases in `test_known_values.py`. - Node tests: scalar factories + new pair-factory block in `bindings/node/__tests__/indicators.test.js`. - Benches: 5 Family-15 benches added in `crates/wickra/benches/indicators.rs`. - Docs: README family-table row + counter (71 -> 88), CHANGELOG entry under [Unreleased]. Note: Family 12 (statistik-regression, PR #51) introduces `node_pair_indicator!` and `wasm_pair_indicator!` macros for Pearson / Beta / Spearman. Family 15 needs the same pair-input pattern but Family 12 is not yet in main, so the three pair wrappers below are written by hand in this PR. When PR #51 lands, the trivial merge-conflict is resolved by keeping the macros from Family 12 and re-using them for Treynor / IR / Alpha (drop the three handwritten wrappers). cargo check --workspace --all-features: green. * fix(family-15): satisfy clippy doc_markdown / if_not_else / digit_grouping * fix(family-15): unused TreynorRatio import, duplicate pairFactories, _eq_nan inf handling * fix(family-15): node eq() handles matching infinities for ratio indicators * test(family-15): cover cold paths flagged by codecov patch |
||
|
kingchenc
|
55284a3042 |
feat(family-14): add 15 candlestick patterns (#53)
* feat(family-14): add 15 candlestick patterns Introduces the Candlestick Patterns family (block A of the family-14 spec) as scalar f64 indicators on Candle inputs. Each detector emits +1.0 for a bullish reading, -1.0 for a bearish reading, and 0.0 when no pattern is present. Doji is direction-less and emits +1.0 / 0.0 only. New indicators (15): - Doji - Hammer - InvertedHammer - HangingMan - ShootingStar - Engulfing - Harami - MorningEveningStar (signed: +1.0 morning star, -1.0 evening star) - ThreeSoldiersOrCrows (signed: +1.0 soldiers, -1.0 crows) - PiercingDarkCloud (signed: +1.0 piercing, -1.0 dark cloud) - Marubozu (signed: +1.0 bullish, -1.0 bearish, 5 percent shadow tolerance default) - Tweezer (signed: +1.0 bottom, -1.0 top, 10 bps relative tolerance default) - SpinningTop (direction-signed indecision) - ThreeInside (confirmed Harami) - ThreeOutside (confirmed Engulfing) MVP scope notes: - Pattern-shape check only, no trend filter applied. Caller combines with a trend indicator for actionable signals. Documented in every doc comment. - Block B (Harmonic patterns) and block C (Chart patterns) remain out-of-scope and will follow when the pattern-detection framework (pivot detector, multi-bar state machines) lands. Touched across all bindings: Python, Node, WASM. Fuzz target, Python tests (streaming-vs-batch + reference values), Node tests (streaming-vs-batch + reference values), and a representative bench subset (1-, 2- and 3-bar patterns) added. README family table + indicator counter (71 -> 86, eight -> nine families) and CHANGELOG [Unreleased] updated. * fix(family-14): unpack MULTI values with *_ to handle 3-element tuples * cov(family-14): cover Default impl cold paths and MorningEveningStar guard branches |
||
|
kingchenc
|
9b8e1346ed |
feat(family-16): add ValueArea + InitialBalance + OpeningRange (#52)
* feat(family-16): add ValueArea + InitialBalance + OpeningRange Opens family #16 (Market Profile) with the three OHLCV-compatible scalar / multi-output indicators: - ValueArea(period, bin_count, value_area_pct) -> {poc, vah, val}. Rolling bin-approximation volume profile over the last `period` candles. Each candle's volume is spread uniformly across [low, high]; POC is the bin with highest cumulative volume; the value area expands symmetrically from POC and always absorbs the higher-volume neighbour next, until `value_area_pct` (default 0.70) of total volume is enclosed. Defaults (20, 50, 0.70). - InitialBalance(period) -> {high, low}. Tracks session-opening high and low over the first `period` bars, then locks. Default period = 12 (one-hour IB on 5-minute bars for US equities). Callers MUST invoke reset() at every session boundary, otherwise IB stays fixed for the lifetime of the instance. - OpeningRange(period) -> {high, low, breakout_distance}. Same lock-after-N-bars semantics as IB with a shorter default period (6 = 30 min on 5-minute bars) and a third output that tracks close - or_mid (positive above the range mid, negative below). Histogram-output Market Profile variants (Volume Profile, VPVR, Composite Profile) are deferred because they need a new histogram output API layer rather than fixed-arity scalars. Tick-data-only variants (TPO Profile, Single Print, Order Flow Delta, Cumulative Delta, Volume-Weighted Open) are out of scope because `wickra-data` does not currently expose tick / L2 data. All four bindings (Rust core, Python, Node, WASM) ship the new indicators with parity tests; benches added; fuzz target extended. Counter 71 -> 74 across 8 -> 9 families. cargo check --workspace --all-features green. * fix(family-16): cover cold paths in InitialBalance + ValueArea InitialBalance::value() public getter had no test covering the post-update Some(...) branch — extended accessors_and_metadata to call value() after one update. ValueArea single-print bar path (c.high == c.low) was unreachable in existing tests since the only single-print test used a uniform 100-price window which exits early via the span == 0 guard; added a mixed-window test that triggers the c.high <= c.low branch directly. The (None, None) arm of the expansion match was by-construction unreachable (the loop condition already requires at least one neighbour) and has been folded into an if/else. |
||
|
kingchenc
|
05fcdd9a5e |
feat(family-12): add 13 Statistik/Regression indicators (#51)
* feat(family-12): add 13 Statistik/Regression indicators Brings the Price Statistics family to 20 indicators (7 → 20) and the total catalogue to 84 (71 → 84). Every indicator ships in the Rust core plus Python, Node, and WASM bindings with full streaming ↔ batch parity, fuzz coverage, and benches. Scalar (f64 → f64): - Variance, CoefficientOfVariation: rolling population variance and its dimensionless ratio with the mean. O(1) updates. - Skewness, Kurtosis: rolling Pearson skewness and excess kurtosis, derived from running sums of x, x², x³, x⁴ via the binomial identities — also O(1) per bar. - StandardError, DetrendedStdDev: standard error of estimate (n − 2) and population StdDev (n) of OLS residuals, sharing the LinReg O(1) sliding sums. - RSquared: coefficient of determination of the rolling OLS fit; the trend-quality filter, clamped to [0, 1]. - MedianAbsoluteDeviation: robust dispersion estimator; O(period log period) per emission via two in-place sorts of a reusable scratch buffer. - Autocorrelation(period, lag): rolling lag-k Pearson autocorrelation. - HurstExponent(period, chunks): R/S-analysis trend-persistence estimator clamped to [0, 1]. Pair indicators (Input = (f64, f64)): - PearsonCorrelation: rolling cross-series Pearson, O(1). - Beta: rolling OLS slope of asset vs. benchmark (CAPM). - SpearmanCorrelation: rolling rank correlation with mid-rank tie handling; O(period log period). Touchpoints: - crates/wickra-core: 13 new indicator modules + mod.rs / lib.rs re-exports. - bindings/python: pyclasses + add_class registration + __init__.py import & __all__ updates. The pair indicators expose update(x, y) and batch(x, y) over two equally-sized numpy arrays. - bindings/node: scalar indicators via node_scalar_indicator! macro; pair indicators via new node_pair_indicator! macro; explicit structs for Autocorrelation and HurstExponent (two-arg ctors). index.js extended with the new exports. - bindings/wasm: scalar wrappers via wasm_scalar_indicator!; pair wrappers via new wasm_pair_indicator! macro. - fuzz: every scalar drove through the generic helper; pair indicators stress-tested by pairing adjacent samples of the fuzz input. - Python tests (test_new_indicators.py): added to SCALAR parametrisation, plus algebraic reference values (variance of [2,4,6] = 8/3, MAD ignoring outlier = 0, monotone non-linear Spearman = 1, two-to-one Beta = 2, etc.) and a streaming-vs-batch test for the pair indicators. - Node tests (indicators.test.js): extended the scalar factories map and added a pair-indicator section with the same algebraic reference values. - crates/wickra/benches: bench_scalar entries for all 10 single- input new indicators. - README: counter 71 → 84; Price Statistics family-table row expanded with the 13 new indicators. - CHANGELOG: Unreleased section documents the family addition. Wiki drafts (ghost-ignored, manual sync to wickra.wiki at release time): indicator-ideas/families/wiki/family-12-statistik-regression/ contains 13 deep-dive pages plus _Sidebar / Indicators-Overview / Warmup-Periods / Home fragments for the curator merge. cargo check --workspace --all-features: clean. * fix(family-12): remove unreachable defensive guards in hurst_exponent The three guards (m < 2 continue, end > buf.len() break, denom == 0.0 return) are by-construction unreachable given the constructor invariant period >= 2 * chunks: m = period / k for k in 1..=chunks always satisfies m >= 2 and end = (c+1) * m <= k * m <= period = buf.len(), and m_1 = period and m_2 = period / 2 are always distinct so the slope denominator is strictly positive. Removing them brings codecov/patch back to 100%. |
||
|
kingchenc
|
5aa0949bce |
feat(family-13): add Ichimoku + Heikin-Ashi (#50)
Two new indicators in a brand-new "Ichimoku & alternative charts" family: - `Ichimoku` (Ichimoku Kinko Hyo): the full five-line cloud system (Tenkan-sen, Kijun-sen, Senkou Span A/B, Chikou Span). Classic (9, 26, 52, 26) defaults; configurable. Forward displacement is handled in an O(1) ring buffer so the visible Senkou A/B at bar n are the values computed at bar n-displacement. - `HeikinAshi`: recursive candle smoothing transform emitting a four-field synthetic candle. Seeds ha_open from (open+close)/2 on the first bar. Touchpoints: core + unit tests, mod.rs/lib.rs re-exports, Python + Node + WASM bindings (multi-output via PyArray2 / interleaved Vec<f64> / Object+Float64Array), Python tests across smoke/new-indicators/ input-validation, Node parity tests, fuzz target (Candle), benches, README family table + counter (71 -> 73, 8 -> 9 families), CHANGELOG. Note: Renko, Kagi, and Point & Figure from the family-13 ideas list are intentionally skipped. They are bar generators (the bar boundary is defined by price moves, not by a fixed time interval) rather than indicators that consume a candle stream, and belong in wickra-data as candle/tick transforms alongside the existing tick-to-candle aggregator and resampler. |
||
|
kingchenc
|
b971e671b4 |
test(family-10): cover cold paths flagged by codecov (#57)
Add tests that exercise the protective fallbacks reachable via flat or
zero-valued input:
- `CenterOfGravity::zero_window_uses_zero_fallback` — den == 0 branch.
- `EhlersStochastic::flat_window_emits_zero` — range == 0 branch.
- `Mama::flat_input_uses_phase_fallback` and the matching
`SineWave` variant — `i1` collapses to zero on a constant series.
- `Fama::new_with_valid_limits_constructs_via_mama` exercises the
`Ok(Self { inner: Mama::new(..)? })` arm that no other test reaches.
Three branches were genuinely unreachable by construction, so the dead
code is removed rather than masked with an attribute:
- `Mama` clamped `alpha > fast_limit` after the lower-bound clamp; the
upper bound is implied by `delta_phase >= 1` and `alpha = fast / delta_phase`.
- `CyberneticCycle` had a `0.0` fallback after the warmup gate that the
3-slot ring buffers preclude (`count >= 7` => all five `Some`s).
- `DecyclerOscillator` used a `let-else { return None }` over a pair of
`Decycler::update` calls that always emit `Some` from the first bar.
|
||
|
kingchenc
|
7a18a26daf |
feat(family-10): add 16 Ehlers / Cycle (DSP) indicators (#49)
Implements Family 10 (Ehlers / Cycle) end-to-end across Rust core,
Python / Node / WASM bindings, fuzz, tests, benches and docs. This
is an entirely new family covering John Ehlers' digital-signal-
processing school of cycle analytics — a strong differentiator
versus TA-Lib and pandas-ta, which ship only fragments.
Indicators:
- MAMA (Mesa Adaptive MA) — multi-output { mama, fama }
- FAMA (Following Adaptive MA) — scalar wrapper around MAMA's slow line
- Fisher Transform — Gaussian-normalising price transform
- Inverse Fisher Transform — bounded oscillator (tanh-based)
- SuperSmoother — 2-pole Butterworth lowpass
- Roofing Filter — high-pass + SuperSmoother bandpass
- Decycler — price minus 2-pole high-pass (lag-free trend)
- Decycler Oscillator — fast / slow Decycler difference (MACD-like)
- Hilbert Dominant Cycle — phase-derived period estimator [6, 50]
- Sine Wave Indicator — sin(phase) with 45° lead companion
- Adaptive Cycle Indicator — half-period driver for adaptive oscillators
- Center of Gravity Oscillator — weighted-mass momentum
- Cybernetic Cycle Component — EasyLanguage classic
- Empirical Mode Decomposition — bandpass + envelope mean
- Ehlers Stochastic — Stochastic on Roofing Filter input, [-1, +1]
- Instantaneous Trendline — Ehlers 2-pole lag-free trend
Indicator count rises 71 -> 87 across nine families (was eight).
All sixteen pass batch == streaming equivalence, expose the standard
Indicator surface (update / batch / reset / is_ready / warmup_period
/ name), are fuzz-tested, benchmarked against the checked-in BTCUSDT
1-minute dataset and reach across all four bindings.
Wiki deep-dive drafts for every indicator + Sidebar / Overview /
Home / Warmup updates are staged under indicator-ideas/families/
wiki/family-10-ehlers-cycle/ in the main repo (ghost-ignored) for
the maintainer to publish to the wiki repo manually.
|
||
|
kingchenc
|
4f9ed34884 |
feat(family-11): add DeMark suite (TD Setup, Sequential, DeMarker, REI, Pressure) (#48)
* feat(family-11): add DeMark suite (TD Setup, Sequential, DeMarker, REI, Pressure)
Family 11 (DeMark) was previously empty; this PR adds five
streaming-first DeMark indicators in one batch.
- **TD Setup** (`TdSetup`): parameterised buy/sell setup counter.
Counts consecutive bars whose close is less-than (buy) or
greater-than (sell) the close `lookback` bars earlier, saturating
at `target`. Emits a signed `f64` so callers read direction from
the sign and run length from the magnitude. Classic config:
`lookback = 4`, `target = 9`.
- **TD Sequential** (`TdSequential`): the canonical Setup + Countdown
exhaustion pattern. Output struct `{ setup, countdown, direction }`
exposes both phase counts as signed numbers plus the active
countdown direction (+1 buy / -1 sell / 0 none). Countdown
activates when a setup completes and tracks the close-vs-high/low
comparison `countdown_lookback` bars back, capped at
`countdown_target`. Classic: 4/9/2/13.
- **TD DeMarker** (`TdDeMarker`): bounded [0, 1] oscillator from the
rolling average of upward high expansion (DeMax) and downward low
expansion (DeMin). Falls back to the neutral 0.5 on a flat market
(denominator zero).
- **TD REI** (`TdRei`): Range Expansion Index, bounded [-100, 100].
Per-bar numerator gated on a range-overlap condition vs the bars
5 and 6 back, normalised by a `period`-bar sum of absolute moves.
Classic period = 5. Saturates at +100 in a slow steady uptrend
and at -100 in the mirror downtrend; emits 0 on a flat market.
- **TD Pressure** (`TdPressure`): volume-weighted buying / selling
pressure normalised to [-100, 100]. Per-bar pressure is the
intra-bar close-vs-open ratio scaled by volume; the output is the
rolling mean divided by the rolling mean volume. Zero-range bars
contribute zero (avoid the undefined ratio) and a flat zero-volume
window falls back to 0.
Bindings: all five exposed in Python (`ta.TDSetup`, `ta.TDSequential`,
`ta.TDDeMarker`, `ta.TDREI`, `ta.TDPressure`), Node (`wickra.TDSetup`
etc.), and WASM. Multi-output classes (`TDSequential`) return either
a struct `{ setup, countdown, direction }` per bar (streaming) or a
flat interleaved Float64Array of length `3 * n` (batch).
Tests: 47 unit tests across the five new core files (pure-trend
saturation, flat-market neutral fallback, batch-equals-streaming,
zero-parameter rejection, reset semantics, accessors). Python
test_new_indicators.py picks up all five plus a multi-output TD
Sequential block. Node indicators.test.js picks up all five.
Reference values added to test_known_values.py.
Fuzz: candle fuzz target sweeps all five DeMark indicators with the
existing `Vec<f64>` -> `Vec<Candle>` driver.
Benches: BTCUSDT 1-minute dataset benches for each DeMark indicator
in `crates/wickra/benches/indicators.rs`.
Docs: README family table gains a "DeMark" row; indicator counter
bumped 71 -> 76. CHANGELOG entry added under [Unreleased]. Wiki
drafts (deep-dive pages + Sidebar / Overview / Warmup-Periods / Home
deltas) live under `indicator-ideas/families/wiki/family-11-demark/`
for manual merge into the wiki repo.
* feat(family-11): add 7 missing DeMark indicators
Complete the DeMark suite (family 11) with the seven indicators not
covered by the first commit: TD Combo, TD Countdown, TD Lines (TDST),
TD Range Projection, TD Differential, TD Open, and TD Risk Level.
- TdCombo: aggressive countdown variant with three strictness rules
on top of the classic close-vs-low/high lookback rule (monotone
low/high, monotone close vs prior bar).
- TdCountdown: standalone 13-bar countdown packaging only the signed
countdown count (the setup machine runs internally).
- TdLines: TDST horizontal support/resistance levels from the
highest-high / lowest-low bars of the most-recently-completed
setup, exposed as a multi-output struct.
- TdRangeProjection: DeMark X-projection of the next bar's high and
low from the current bar's OHLC via an open-vs-close-weighted
pivot (three branches: close<open, close>open, close==open).
- TdDifferential: two-bar buying-pressure vs selling-pressure
reversal pattern emitting +1/-1/0.
- TdOpen: gap-and-fade reversal pattern (open outside prior range
with subsequent recovery into it) emitting +1/-1/0.
- TdRiskLevel: protective stop levels derived from the setup
extreme bar +/- its true range.
All seven are wired through Rust core, Python, Node and WASM
bindings, registered in the candle-stream fuzz target, given
benchmark entries on the BTCUSDT 1-minute dataset, and covered by
streaming-vs-batch equivalence, reference-value, lifecycle and
input-validation tests on the Python and Node sides. README counter
moves 76 -> 83 and the CHANGELOG "family 11" entry is extended to
list all twelve indicators.
* fix(td_risk_level tests): check first emission at idx 12, not last bar
TdRiskLevel re-ratchets the sell-risk level on each subsequent setup
completion, so a strictly rising series produces 22.0 at idx 19 (latest
setup) rather than 15.0 (first setup). The test comment already named
idx 12 as the reference; switch the assertion from out[-1] to out[12]
to match the reference computation.
* test(family-11): cover buy-direction branches in TD indicators
Add downtrend tests to TdSequential, TdCombo and TdCountdown so the
buy-side countdown/combo increment branches are exercised; remove an
empty `if buy_countdown == target {}` block in TdSequential whose
behavior is already enforced by the outer strict `<` guard.
Closes codecov/patch gaps reported on PR #48 (10 missed lines across
the three files).
|
||
|
kingchenc
|
7e1e988596 |
feat(family-08): Pivots & Support/Resistance (7 indicators) (#47)
* feat(family-08): add Classic, Fibonacci, Camarilla, Woodie and DeMark pivots + Williams Fractals + ZigZag
Seven new indicators land the previously empty Pivots & S/R family
(family 08), each implemented in wickra-core with the full Indicator
trait surface (update / reset / warmup_period / is_ready / name),
exposed across Python (PyO3), Node (napi-rs) and WASM (wasm-bindgen)
with the standard streaming + batch APIs, and covered by Rust unit
tests, Python streaming-vs-batch + reference-value tests, Node
streaming-vs-batch tests, the candle-input fuzz target and Rust
microbenchmarks.
- ClassicPivots (7 levels): PP = (H+L+C)/3, three R/S tiers per the
floor-trader formulas.
- FibonacciPivots (7 levels): PP plus R/S spaced by 0.382 / 0.618 /
1.000 of the prior range.
- Camarilla (9 levels): Nick Stott's four-tier `C +/- (H - L) * 1.1 /
{12, 6, 4, 2}` levels.
- WoodiePivots (5 levels): close-weighted PP = (H + L + 2*C) / 4 plus
two R/S tiers.
- DemarkPivots (3 levels): conditional X sum based on the previous
bar's open-vs-close relationship.
- WilliamsFractals: five-bar swing detector emitting optional up/down
fractal prices at the centre of each window.
- ZigZag: percent-threshold swing tracker, non-repainting; emits the
just-completed extreme and direction on confirmed reversals only.
README family table updated to nine families / 78 indicators;
CHANGELOG records the family-08 addition under [Unreleased].
* fix(family-08 tests): unify MULTI dict to 3-tuple (factory, batch_call, k)
The HEAD-side family-08 test parametrised MULTI[name] as
`(factory, batch_call, output_arity)` so that pivots with arity 3/5/7/9
fit the same harness. Main's entries arrived as 2-tuples; convert them
all to the 3-tuple shape so `make, batch_call, k = MULTI[name]` unpacks
cleanly. Lifecycle test now indexes the tuple instead of destructuring.
* test(zig_zag): tighten flat-oscillation test (drop dead counter branch)
The previous version of `small_oscillations_yield_no_swings` counted
emitted swings, but the assertion proves the counter never increments
so codecov flagged `emitted += 1` as uncovered. Switch to a per-bar
`assert!(...is_none())` — same coverage of the no-swing path, no dead
branch.
|
||
|
kingchenc
|
f10b8c2e2d |
feat(family-09): add 7 trailing stops (HiLo, Volty, Yo-Yo, Donchian, Pct, Step, Renko) (#46)
* feat(family-09): add 7 trailing stops (HiLo, Volty, Yo-Yo, Donchian, Pct, Step, Renko)
Rounds out the Trailing Stops family from 5 to 12 indicators:
- HiLoActivator (Crabel): SMA-of-high/SMA-of-low trail with a one-bar
lag; emits the opposite-side SMA as the trailing stop.
- VoltyStop (Cynthia Kase): ATR trail anchored on the extreme close
since the trade was opened — tighter than AtrTrailingStop on
pullbacks.
- YoyoExit: long-only ATR trail with an explicit re-entry trigger at
trail + multiplier*ATR; exposes an in_trade flag.
- DonchianStop (Turtle): lowest low / highest high over the window;
multi-output {stop_long, stop_short}.
- PercentageTrailingStop: fixed-percent trail that scales across
instruments without per-asset tuning.
- StepTrailingStop: snaps to a step_size-aligned grid; mirrors
discretionary stop-by-hand workflow.
- RenkoTrailingStop: block-anchored trail; only moves on full-block
advances, ignores intra-block noise.
All seven are wired into wickra-core, the Python / Node / WASM
bindings, the indicator_update + indicator_update_candle fuzz targets,
the wickra bench harness, and the Python + Node test suites. README
counter bumps from 71 to 78; CHANGELOG entry under [Unreleased].
* fix(family-09): satisfy pedantic clippy lints
- hilo_activator: rewrite match-Some/None as if-let-else (single_match_else),
add backticks around the HiLo identifier in module/struct doc (doc_markdown).
- percentage / step / renko trailing stop tests: use f64::from(i32) instead
of `as f64` (cast_lossless).
- bench `benches()` is now >100 lines after Family 09 was wired in; allow
too_many_lines (matches the python pymodule fn).
|
||
|
kingchenc
|
880a0e7430 |
feat: Family 07 Volume - 6 new volume-flow indicators (#45)
* feat(kvo): add Klinger Volume Oscillator
Stephen J. Klinger's trend-aware volume-force MACD. Each bar produces a 'volume force' (vf) signed by the local trend (+1 / -1 / carry) and scaled by the ratio of the current accumulation horizon to its previous trend. KVO = EMA(vf, fast) - EMA(vf, slow), classic (34, 55).
Rust core (Kvo) with 7 unit tests (rejects zero / fast>=slow, accessors, constant series collapses to 0, warmup lands at slow+1, batch == streaming, reset clears state), plus Python (PyKvo + KVO export), Node (KvoNode), and WASM (WasmKvo) bindings. Fuzz target adds Kvo to the candle-input sweep, bench adds the candle-input KVO benchmark, README counter 71 -> 72 + family table row, CHANGELOG [Unreleased].
* feat(volume-oscillator): add Volume Oscillator (VO)
Percent difference between a fast and a slow SMA of the bar volume: 100 * (SMA(vol, fast) - SMA(vol, slow)) / SMA(vol, slow). Default (14, 28). The line stays near zero in stable conditions; positive readings show rising short-term participation, negative readings show waning interest.
Rust core (VolumeOscillator) with 8 unit tests (period validation, accessors, constant volume == 0, zero-volume window defensive branch, two reference values verified algebraically, batch == streaming, reset), plus Python (PyVolumeOscillator + VolumeOscillator export), Node (VolumeOscillatorNode), and WASM (WasmVolumeOscillator) bindings. Fuzz target adds VolumeOscillator to the candle-input sweep, bench adds the volume_oscillator benchmark, README counter 72 -> 73 + family table row, CHANGELOG [Unreleased].
* feat(nvi-pvi): add Negative & Positive Volume Index
Paul Dysart's cumulative volume-flow indices, popularised by Norman Fosback in 'Stock Market Logic'. Both run from a 1000.0 baseline and only update on a specific direction of volume change:
- NVI updates on volume-contraction bars (volume_t < volume_{t-1}), absorbing the percent close change. Tracks the 'smart money' leg per Fosback.
- PVI updates on volume-expansion bars (volume_t > volume_{t-1}). Tracks the 'crowd' leg.
Both expose with_baseline(f64) for custom starting indexes. The NVI/PVI pair is listed as a single line in indicator-ideas/families/07-volume.md and shares the same lifecycle/test/binding surface, so they ship as one commit.
Rust core (Nvi, Pvi) with 9 unit tests each (accessors, baseline seed, volume direction branches, zero-prev-close guard, custom baseline, batch == streaming, reset), plus Python (PyNvi/PyPvi + NVI/PVI exports), Node (NviNode/PviNode), and WASM (WasmNvi/WasmPvi) bindings. Fuzz target adds Nvi+Pvi to the candle-input sweep, bench adds nvi+pvi entries, README counter 73 -> 75 + family table row, CHANGELOG [Unreleased].
* feat(family-07): add Williams A/D, Anchored VWAP, Demand Index, TSV, VZO, Market Facilitation Index
Finishes the volume-flow family with the remaining (new) entries from
indicator-ideas/families/07-volume.md.
Indicators added:
- Williams A/D (`WilliamsAD`): Larry Williams' volume-less cumulative
accumulation/distribution line. Anchors each bar's contribution to
the previous close via true-high/true-low (gap-aware).
- Anchored VWAP (`AnchoredVwap`): cumulative VWAP whose accumulation
starts at a user-chosen anchor bar. Exposes `set_anchor()` (queued
to the next `update`) for click-to-anchor workflows. Reset clears
both state and pending-anchor flag.
- Demand Index (`DemandIndex`): James Sibbet's smoothed buying-vs-
selling pressure, in the streaming-friendly textbook form
`EMA(volume * close-return * (1 + range/close), period)`.
- Time Segmented Volume (`Tsv`): Don Worden's rolling window-sum of
`(close_t - close_{t-1}) * volume_t`. Default `period = 18`.
- Volume Zone Oscillator (`Vzo`): Walid Khalil's normalised volume-flow
oscillator bounded in `[-100, +100]`, defined as
`100 * EMA(signed_volume) / EMA(volume)`.
- Market Facilitation Index (`MarketFacilitationIndex`): Bill Williams'
per-bar `(high - low) / volume`. Returns `None` on zero-volume bars.
All six indicators ship with unit tests (`rejects_zero_period` where
applicable, `accessors_and_metadata`, constant-series behaviour,
batch == streaming equivalence, reset semantics, and reference-value
or saturation-extreme tests), Python / Node / WASM bindings, fuzz
coverage in `indicator_update_candle`, a `bench_candle_input` line per
indicator, README + CHANGELOG entries, and Python reference-value
tests in `test_new_indicators.py`.
The README indicator counter advances 75 -> 81.
* test(family-07): cover defensive cold paths + Default impls
- ad_oscillator: exercise `value()` after first emission.
- kvo: cover the `cm == 0.0` zero-OHLC defensive branch.
- nvi / pvi: exercise the Default impls.
|
||
|
kingchenc
|
6287bd48c1 |
feat: Family 06 Trend-Strength - 5 new directional/random-walk indicators (#44)
* feat(adxr): add Wilder Average Directional Movement Index Rating
ADXR is the trend-strength smoother Wilder published alongside ADX in
*New Concepts in Technical Trading Systems* (1978):
ADXR_t = (ADX_t + ADX_{t - (period - 1)}) / 2
The lookback length is the same period that feeds the underlying ADX.
Because the older ADX is period - 1 bars stale, ADXR responds more
slowly than ADX and is the canonical metric for comparing
trend-strength across instruments.
Implementation reuses the existing wickra_core::Adx engine plus a
period-length ring of past ADX values; warmup is 3 * period - 1
(41 for period = 14). Bindings: Python PyAdxr (PyArray1 batch),
Node AdxrNode (number scalar), WASM WasmAdxr. Fuzz target covers
the candle-input path. Python + Node streaming-vs-batch tests
parametrised, plus a pure-uptrend reference value (ADXR == 100
when ADX saturates at 100). Criterion bench added under crates/
wickra/benches/indicators.rs.
README family table and indicator counter updated (71 -> 72).
* feat(rwi): add Mike Poulos Random Walk Index
RWI compares actual price displacement to what a random walk would
produce over the same horizon: for each lookback i in [2, period],
RWI_High_t(i) = (high_t - low_{t-i+1}) / (ATR_i(t) * sqrt(i))
RWI_Low_t(i) = (high_{t-i+1} - low_t) / (ATR_i(t) * sqrt(i))
Per-bar output is the maximum across lookbacks for each direction;
a reading > 1 means the trend beats random-walk noise, > 2 is the
typical strong-trend threshold. Multi-output (high, low). period
must be >= 2 (the shortest meaningful lookback); period < 2 returns
InvalidPeriod. Warmup = period (e.g. 14 for the standard default).
Bindings: Python PyRwi (PyArray2 shape (n, 2)), Node RwiNode +
RwiValue struct, WASM WasmRwi (Object/Reflect for update,
Float64Array interleaved for batch). Fuzz target adds the candle
input case. Python parametric streaming-vs-batch test and pure
uptrend reference test (RWI_High dominates RWI_Low and exceeds 1).
Node parametric streaming-vs-interleaved-batch test. Criterion
bench under crates/wickra/benches/indicators.rs.
README family table and indicator counter updated (72 -> 73).
* feat(tii): add M.H. Pee Trend Intensity Index
TII is a [0, 100] oscillator that asks 'what fraction of the recent
SMA deviations are positive?'. The construction is
dev_t = close_t - SMA(close, sma_period)_t
SD_pos = sum of positive dev_t over the last dev_period bars
SD_neg = sum of |negative dev_t| over the last dev_period bars
TII = 100 * SD_pos / (SD_pos + SD_neg)
Saturates at 100 on a pure uptrend (every close above the lagging
SMA), at 0 on a pure downtrend, and returns the neutral mid-point 50
on a perfectly flat window. The output is clamped to [0, 100] as
the rolling-sum subtraction loop can accumulate a few ULP of error
on long histories. Canonical Pee parameters (sma_period=60,
dev_period=30) wired as Python defaults; warmup is
sma_period + dev_period - 1 (89 for the defaults).
Bindings: Python PyTii (PyArray1 batch), Node TiiNode (scalar
update + batch), WASM WasmTii via the two-arg wasm_scalar_indicator!
macro. Fuzz target adds the scalar path. Python parametric
streaming-vs-batch test plus pure-uptrend (TII == 100) and
flat-market (TII == 50) reference tests. Node parametric
streaming-vs-batch test. Criterion bench under crates/wickra/
benches/indicators.rs.
README family table and indicator counter updated (73 -> 74).
* feat(kst): add Pring Know Sure Thing oscillator
KST is Martin Pring's long-horizon momentum gauge: four smoothed
rate-of-change components combined with fixed weights (1, 2, 3, 4),
plus an SMA signal line.
RCMA_i = SMA(ROC(close, roc_i), sma_i) for i in 1..=4
KST = 1*RCMA_1 + 2*RCMA_2 + 3*RCMA_3 + 4*RCMA_4
Signal = SMA(KST, signal_period)
Kst::classic() exposes Pring's recommended parameter set
(roc = (10, 15, 20, 30), sma = (10, 10, 10, 15), signal = 9);
warmup = max(roc_i + sma_i) + signal_period - 1 (53 for the classic
parameters). All four parallel branches are fed unconditionally so
they warm in lock-step.
Bindings: Python PyKst (PyArray2 shape (n, 2)) with a KST.classic()
staticmethod, Node KstNode + KstValue with a KST.classic() factory,
WASM WasmKst with both new(...) and classic() constructors plus
Object/Reflect for update and Float64Array for batch. Fuzz target
adds the scalar multi-output path. Python tests gain a new
MULTI_SCALAR section parametric over scalar-input/multi-output
indicators, plus a classic-on-constant-series reference test. Node
tests gain a KST entry in the multi-output section. Criterion
benchmark added under crates/wickra/benches/indicators.rs.
README family table and indicator counter updated (74 -> 75).
* feat(wave-trend): add LazyBear Wave Trend Oscillator
Two-line mean-reverting momentum gauge built from the typical price
and three cascaded EMAs:
ap = (high + low + close) / 3
esa = EMA(ap, channel_period)
d = EMA(|ap - esa|, channel_period)
ci = (ap - esa) / (0.015 * d)
wt1 = EMA(ci, average_period)
wt2 = SMA(wt1, signal_period)
WaveTrend::classic() exposes LazyBear's defaults
(channel = 10, average = 21, signal = 4); warmup is
2 * channel_period + average_period + signal_period - 3 (42 for the
classic defaults). On a perfectly flat market the SMA-seeded EMA
introduces a single-ULP drift between ap and esa, which on a tiny d
would make the ratio explode to -1/0.015 = -66.67; a price-scaled
flat-tolerance guard (d <= 16 * EPSILON * max(|esa|, 1)) collapses
the channel index to 0 in that regime so both lines remain at zero.
Bindings: Python PyWaveTrend (PyArray2 shape (n, 2)) with a
WaveTrend.classic() staticmethod, Node WaveTrendNode + WaveTrendValue
with a WaveTrend.classic() factory, WASM WasmWaveTrend with both
new(...) and classic() constructors. Fuzz target adds the candle
multi-output path (sorted alphabetically). Python parametric
streaming-vs-batch test plus a flat-market reference test. Node
parametric streaming-vs-interleaved-batch test. Criterion bench
under crates/wickra/benches/indicators.rs.
README family table and indicator counter updated (75 -> 76).
* fix(family-06): re-add KST::classic() factory + drop dup fuzz block
Family-06 PR's tests call ta.KST.classic() / wickra.KST.classic() — main's
KST binding shipped without the static factory. Add classic() in Python
(staticmethod) and Node (napi factory); WASM already had it. Also drop the
duplicate Kst::classic().unwrap() block in fuzz/indicator_update.rs that
the merge left behind (main's API no longer returns Result).
* test(rwi): drop dead count==0 guard
The loop `for i in 2..=period` makes `count = tr_end - tr_start = i - 1`
which is always >= 1, so the `if count == 0 { continue; }` branch was
unreachable defensive code that codecov flagged on the family-06 PR.
|
||
|
kingchenc
|
54194a4ff8 |
feat: Family 05 Bands & Channels - 11 new price-envelope indicators (#43)
* feat(bands-channels): add Family 05 with 11 indicators
Eleven price-envelope overlays organised into a new "Bands & Channels"
family, exposed across all four bindings (Rust core, Python, Node, WASM)
plus fuzz/test/bench/docs coverage:
- MaEnvelope - SMA centerline with fixed-percent envelope (the oldest
band overlay still in regular use).
- AccelerationBands (Price Headley) - momentum-biased bands that widen
with the bar's relative range (H - L) / (H + L).
- StarcBands (Stoller Average Range Channel) - SMA(close) +/- k*ATR;
Keltner's SMA-centerline sibling.
- AtrBands - close-anchored envelope of width k*ATR; the standard
volatility-targeting stop/target band.
- HurstChannel - SMA centerline wrapped by the rolling high-low range
(Brian Millard / Hurst-cycle channel).
- LinRegChannel - rolling OLS endpoint +/- k * population stddev of the
residuals; dispersion about the trend rather than the mean.
- StandardErrorBands - regression line +/- k * OLS standard error
(denominator n - 2) for prediction-interval bands.
- DoubleBollinger (Kathy Lien) - two concentric BB envelopes
(typically +/- 1 sigma and +/- 2 sigma) for the zone-partition setup.
- TtmSqueeze (John Carter) - BB-inside-KC squeeze flag paired with a
detrended-close linear-regression momentum reading.
- FractalChaosBands - Bill Williams 5-bar fractal high/low envelope.
- VwapStdDevBands - cumulative VWAP with volume-weighted population
standard deviation bands.
Each indicator ships:
- Core impl with the full Indicator trait, classic() where applicable,
and unit tests (rejects_zero_period / multiplier, accessors, flat
market, monotonic ordering, batch == streaming, reset, plus
algebraically verifiable reference values).
- Python PyO3 binding with multi-column NumPy batch (PyArray2).
- Node napi binding with #[napi(object)] struct + interleaved flat
batch.
- WASM wasm-bindgen binding via Object/Reflect for update +
Float64Array for batch.
- Fuzz coverage in fuzz_targets/indicator_update{,_candle}.rs.
- Python streaming-vs-batch parametric test + reference test.
- Node streaming-vs-interleaved-batch test + reference test.
- Criterion microbench under crates/wickra/benches/indicators.rs.
README family table, README indicator-count line, and CHANGELOG
Unreleased entry updated: indicator total rises from 71 to 82 across
nine families. Wiki pages are updated in a separate commit in the
wickra.wiki repo.
* test(acceleration-bands): cover sum_hl==0 zero-price guard
Exercises line 104 (`0.0` branch of the `sum_hl == 0.0` guard) which
was the last patch-coverage miss on the family-05 PR. `Candle::new`
accepts a fully-zero bar so the branch is reachable in principle —
add a degenerate-candle unit test to hit it.
|
||
|
kingchenc
|
3ea0f12b7a |
feat: Family 04 Volatility — RVI / Parkinson / Garman-Klass / Rogers-Satchell / Yang-Zhang (#42)
* feat(rvi): add Relative Volatility Index
Donald Dorsey's RSI-shaped volatility gauge. Partitions the rolling
population standard deviation of close into "up" samples (close rose
since the previous bar) and "down" samples (close fell), Wilder-smooths
each side, and reports 100 * AvgUp / (AvgUp + AvgDown). Output bounded
on [0, 100]; saturates at 100 in pure uptrends, 0 in pure downtrends,
and falls back to 50 on a completely flat series (same undefined-RS
convention as RSI).
Single period parameter (default 10) drives both the stddev window and
the Wilder smoothing constant. First emit lands at index 2*period - 2
(2*period - 1 bars are needed: period to fill the stddev window plus
period - 1 to seed the Wilder averages, overlapping by one bar).
Touchpoints: rvi.rs + mod.rs + lib.rs re-export, PyRvi + __init__.py +
test_new_indicators SCALAR + test_known_values uptrend reference,
RviNode + index.d.ts/index.js + indicators.test.js factory +
reference, WasmRvi via scalar macro, scalar-fuzz target, bench_scalar
entry, README + CHANGELOG.
* feat(parkinson): add Parkinson Volatility
Michael Parkinson's (1980) high-low realised volatility estimator.
Under a driftless Geometric-Brownian-Motion assumption, the extreme
range of a bar carries roughly 5x the variance information of the
close-to-close estimator, so for a given statistical efficiency
Parkinson needs five times fewer samples.
Formula:
sigma^2 = (1 / (4n * ln 2)) * Sum_{i=1..n} (ln(H_i / L_i))^2
out = sqrt(sigma^2) * sqrt(trading_periods) * 100
The output is annualised to a percent in the same style as
HistoricalVolatility (pass `trading_periods = 1` for the raw per-bar
sigma * 100 figure). Two parameters: `period` (default 20) for the
rolling window, `trading_periods` (default 252) for the annualisation
factor. First emit at index `period - 1`.
Touchpoints: parkinson.rs + mod.rs + lib.rs re-export,
PyParkinsonVolatility + __init__.py + test_new_indicators CANDLE_SCALAR
+ test_known_values zero-range reference, ParkinsonVolatilityNode +
index.d.ts/index.js + indicators.test.js factory + reference,
WasmParkinsonVolatility hand-rolled, candle-fuzz target,
bench_candle_input entry, README + CHANGELOG.
* feat(garman-klass): add Garman-Klass Volatility
Garman & Klass (1980) OHLC realised-volatility estimator. Extends
Parkinson's high-low estimator with an open-to-close term, lifting
statistical efficiency from ~5x to ~7.4x relative to close-to-close
stddev under driftless Geometric Brownian Motion.
Formula (per bar):
s_t = 0.5 * (ln(H_t / L_t))^2 - (2*ln(2) - 1) * (ln(C_t / O_t))^2
out = sqrt(max(mean(s_t over `period`), 0)) * sqrt(trading_periods) * 100
The per-bar sample can be marginally negative when the bar has a small
range relative to its open-to-close move; a max(., 0) clamp on the
rolling mean absorbs that and the FP cancellation noise before the
square root.
Still biased on data with meaningful overnight drift -- use Yang-Zhang
when gaps matter. Defaults: `period = 20`, `trading_periods = 252`
(annualised percent, same convention as HistoricalVolatility).
Touchpoints: garman_klass.rs + mod.rs + lib.rs re-export,
PyGarmanKlassVolatility + __init__.py + test_new_indicators
CANDLE_SCALAR + test_known_values zero-movement reference,
GarmanKlassVolatilityNode + index.d.ts/index.js + indicators.test.js
factory + reference, WasmGarmanKlassVolatility hand-rolled,
candle-fuzz target, bench_candle_input entry, README + CHANGELOG.
* feat(rogers-satchell): add Rogers-Satchell Volatility
Rogers, Satchell & Yoon (1994) OHLC realised-volatility estimator.
Unlike Garman-Klass, the per-bar sample is exact under arbitrary
Brownian drift -- the drift component cancels algebraically.
Formula (per bar):
s_t = ln(H_t / C_t) * ln(H_t / O_t) + ln(L_t / C_t) * ln(L_t / O_t)
out = sqrt(max(mean(s_t over `period`), 0)) * sqrt(trading_periods) * 100
Each per-bar sample is also non-negative by construction: with
`Candle::new` guaranteeing H >= max(O, L, C) and L <= min(O, H, C), the
four log factors have predictable signs (ln(H/.) >= 0, ln(L/.) <= 0),
so both products contribute >= 0. The max(., 0) clamp on the rolling
mean is only there to absorb FP cancellation.
Defaults: `period = 20`, `trading_periods = 252` (annualised percent,
same convention as HistoricalVolatility / Parkinson / Garman-Klass).
Touchpoints: rogers_satchell.rs + mod.rs + lib.rs re-export,
PyRogersSatchellVolatility + __init__.py + test_new_indicators
CANDLE_SCALAR + test_known_values zero-movement reference,
RogersSatchellVolatilityNode + index.d.ts/index.js + indicators.test.js
factory + reference, WasmRogersSatchellVolatility hand-rolled,
candle-fuzz target, bench_candle_input entry, README + CHANGELOG.
* feat(yang-zhang): add Yang-Zhang Volatility
Yang & Zhang (2000) drift- and gap-robust OHLC realised-volatility
estimator. Combines three independent components into a single estimate
with minimum variance:
overnight = sample_var(ln(O_t / C_{t-1})) over n bars (close-to-open)
open_close = sample_var(ln(C_t / O_t)) over n bars
rs = mean(ln(H/C)*ln(H/O) + ln(L/C)*ln(L/O)) over n bars
sigma^2_YZ = overnight + k*open_close + (1-k)*rs
k = 0.34 / (1.34 + (n+1)/(n-1))
out = sqrt(max(sigma^2_YZ, 0)) * sqrt(trading_periods) * 100
The overnight and open-to-close variances use Bessel's correction (the
sample estimator, divisor n-1), same convention as
HistoricalVolatility. The blending factor `k` is the one that
minimises estimator variance under driftless Geometric Brownian Motion
with overnight gaps.
This is the gold-standard OHLC estimator for assets with both
close-to-open gaps and intraday drift: equities, futures, and any
market that does not trade continuously. For pure intraday data (where
O_t == C_{t-1} and the open-to-close return is constant), the
overnight and open-close terms vanish and the estimator collapses to
(1-k) * Rogers-Satchell -- this is the indicator's
intraday_data_collapses_to_rs_only unit test.
Period >= 2 (Bessel correction needs >= 2 samples). First emit at
index `period` (the (period+1)-th bar): one bar seeds prev_close, the
next `period` fill the rolling windows. Defaults: `period = 20`,
`trading_periods = 252`.
Touchpoints: yang_zhang.rs + mod.rs + lib.rs re-export,
PyYangZhangVolatility + __init__.py + test_new_indicators
CANDLE_SCALAR + test_known_values zero-movement reference,
YangZhangVolatilityNode + index.d.ts/index.js + indicators.test.js
factory + reference, WasmYangZhangVolatility hand-rolled, candle-fuzz
target, bench_candle_input entry, README + CHANGELOG.
* fix(rvi): rename to RviVolatility to avoid clash with family-02 RVI
Family 02 (PR #40) ships a separate `Rvi` struct for Relative Vigor
Index. The two indicators have nothing to do with each other beyond
sharing the acronym, so disambiguate by giving the volatility one a
longer name everywhere:
- Rust crate: `Rvi` -> `RviVolatility`
- Rust file: `rvi.rs` -> `rvi_volatility.rs`
- Python: `RVI` -> `RVIVolatility`
- Node: `RVI` -> `RVIVolatility`
- WASM: `RVI` -> `RVIVolatility`
Once the two PRs are both merged, callers get `wickra::Rvi` for Vigor
and `wickra::RviVolatility` for Volatility. The shorter `RVI` acronym
stays with the Momentum family per the existing wiki pages and the
implementation that shipped first.
Updates: rvi_volatility.rs (renamed), mod.rs, lib.rs re-export,
bindings/python/src/lib.rs + __init__.py + tests, bindings/node/src/lib.rs
+ index.d.ts + index.js + __tests__, bindings/wasm/src/lib.rs,
fuzz/fuzz_targets/indicator_update.rs, crates/wickra/benches/indicators.rs,
README family-table label, CHANGELOG entry.
* test(volatility): Rename test_rvi -> test_rvi_volatility + drop dead match arms
The Python test test_rvi_pure_uptrend_saturates_at_one_hundred was
calling ta.RVI() expecting the volatility version, but ta.RVI now
means Family 02's Relative Vigor Index (candle input). Renamed to
ta.RVIVolatility to match the binding rename done at merge time.
In all four OHLC volatility tests, the existing `match (r, a) { ...,
_ => panic!() }` arm is dead in passing runs (every aligned pair is
either (None, None) or (Some, Some)). Codecov flagged it as a patch
miss on each of parkinson / garman_klass / rogers_satchell /
yang_zhang. Refactored per CLAUDE.md cold-path guidance to
`assert_eq!(r.is_some(), a.is_some()); if let (Some, Some) ...`.
|
||
|
kingchenc
|
d9d3ad18aa |
feat: Family 03 MACD & Price Oscillators — APO / AO-Hist / CFO / Zero-Lag MACD / Elder Impulse / STC (#41)
* feat(apo): add Absolute Price Oscillator EMA(close, fast) - EMA(close, slow). Like MACD without the signal EMA. Defaults to (fast = 12, slow = 26); fast must be strictly less than slow. Touchpoints: apo.rs + mod.rs + lib.rs re-export, PyApo + __init__.py + test_new_indicators SCALAR + test_known_values flat reference, ApoNode + index.d.ts/index.js + indicators.test.js factory + reference, WasmApo via scalar macro, scalar-fuzz target, README + CHANGELOG. * fix(apo): add PyApo + ApoNode + WasmApo bindings missed from |
||
|
kingchenc
|
7f1a6df202 |
ci(sync-about): Push counter fix to PR branch instead of main (#56)
Previously the workflow patched README.md on main after every push, producing an unsigned 'chore: sync indicator count' commit per merge. Now the README counter is kept in sync on the PR side instead: on every pull_request event, the workflow checks out the PR's head ref, compares grep -c '^mod ' to the README counter, and if they differ, pushes a fix-up commit back onto the PR branch using the default GITHUB_TOKEN. When the PR is squash-merged, that fix-up commit is folded into the single web-flow-signed merge commit on main — so main's history never shows a separate bot commit. About description and Wiki sync still run on push to main / v* tags via the existing PAT, since both reach outside the main repo (Administration:write and the .wiki repo). For PRs from forks the workflow cannot push back; it emits a hard ::error:: pointing at README.md so the contributor can fix the counter manually. Pushes via GITHUB_TOKEN do not re-trigger downstream workflows (GitHub's anti-recursion policy), so the fix-up commit costs zero extra CI minutes — only sync-about itself re-runs on the next synchronize event and no-ops once the counter matches. |
||
|
wickra-bot
|
1ea05fb2a1 | chore: sync indicator count to 85 [skip ci] | ||
|
kingchenc
|
24e723fa7d |
feat: Family 02 Momentum Oscillators — RVI / PGO / KST / SMI / Laguerre / Connors / Inertia (#40)
* feat(rvi): add Relative Vigor Index
Dorsey's RVI = SMA(close - open, period) / SMA(high - low, period) over
a rolling window of period candles. Candle input, single parameter
period (default 10). Positive on average-bullish windows, negative on
average-bearish. Holds the previous value if the entire window has
zero range (denominator undefined).
Reference: Donald Dorsey, also pandas-ta rvi.
Touchpoints: rvi.rs + mod.rs + lib.rs re-export, PyRvi + __init__.py
+ test_new_indicators CANDLE_SCALAR + test_known_values reference,
RviNode (4-column OHLC batch) + index.d.ts/index.js + indicators.test
.js factory + reference, WasmRvi + make_candle_ohlc helper, candle-fuzz
target + criterion bench, README + CHANGELOG.
* feat(pgo): add Pretty Good Oscillator
Mark Johnson's PGO = (close - SMA(close, period)) / EMA(TR, period).
Counts roughly how many ATR-equivalents the close sits from its
period-bar mean. Candle input, single parameter period (default 14).
Johnson's heuristic uses +3/-3 crossings as entry signals.
Touchpoints: pgo.rs + mod.rs + lib.rs re-export, PyPgo + __init__.py
+ test_new_indicators CANDLE_SCALAR + test_known_values flat-close
reference, PgoNode (h/l/c) + index.d.ts/index.js + indicators.test.js
factory + reference, WasmPgo, candle-fuzz target + bench, README +
CHANGELOG.
* feat(kst): add Know Sure Thing (Pring)
Pring's long-horizon momentum oscillator: weighted sum of four
SMA-smoothed ROC series with fixed weights 1, 2, 3, 4, plus an SMA
signal line. Nine parameters (four ROC periods, four SMA periods, one
signal period); classic() applies Pring's recommended defaults.
Multi-output indicator emitting KstOutput { kst, signal }.
Touchpoints: kst.rs + mod.rs + lib.rs re-export, PyKst + __init__.py
+ test_new_indicators MULTI + test_known_values flat-input reference,
KstNode + KstValue + index.d.ts/index.js + indicators.test.js multi
factory + reference, WasmKst (manual JsValue object), scalar-fuzz
target (handled outside the f64-output drive helper), README +
CHANGELOG.
* feat(smi): add Stochastic Momentum Index (Blau)
Blau's doubly-EMA-smoothed bounded oscillator: measures the close's
displacement from the centre of the recent high-low range, scaled by
the smoothed range. Candle input, three parameters (period, d_period,
d2_period) with defaults 5 / 3 / 3.
Internally feeds both the displacement-EMA stack and the range-EMA
stack on every candle so they warm up in parallel (gating either
behind the other starves the second by one input).
Touchpoints: smi.rs + mod.rs + lib.rs re-export, PySmi + __init__.py
+ test_new_indicators CANDLE_SCALAR + test_known_values flat-input
reference, SmiNode + index.d.ts/index.js + indicators.test.js factory
+ reference, WasmSmi, candle-fuzz target, README + CHANGELOG.
* feat(laguerre-rsi): add Ehlers Laguerre RSI
Four-stage Laguerre polynomial filter wrapped in an RSI-style up/down
accumulator. Single gamma in [0, 1] (default 0.5) trades lag for
smoothness. State is seeded by setting all four L_i to the first input
so a constant series stays at the neutral 50. Output clamped to
[0, 100] to absorb floating-point rounding.
Reference: Ehlers, Time Warp - Without Space Travel, 2002.
Touchpoints: laguerre_rsi.rs + mod.rs + lib.rs re-export, PyLaguerreRsi
+ __init__.py + test_new_indicators SCALAR + test_known_values neutral
reference, LaguerreRsiNode + index.d.ts/index.js + indicators.test.js
factory + reference, WasmLaguerreRsi via scalar macro, scalar-fuzz
target, README + CHANGELOG.
* feat(connors-rsi): add Connors RSI (CRSI)
Larry Connors' 3-component aggregate: RSI(close), RSI(streak), and
PercentRank of the 1-period return over the last period_rank returns.
Each component is bounded in [0, 100] so the aggregate is too.
Three parameters (period_rsi, period_streak, period_rank) with
defaults 3 / 2 / 100. Streak tracks consecutive up/down runs (resets
to 0 on unchanged close).
Touchpoints: connors_rsi.rs + mod.rs + lib.rs re-export, PyConnorsRsi
+ __init__.py + test_new_indicators SCALAR + test_known_values bounded
reference, ConnorsRsiNode + index.d.ts/index.js + indicators.test.js
factory + reference, WasmConnorsRsi via scalar macro, scalar-fuzz
target, README + CHANGELOG.
* feat(inertia): add Dorsey Inertia (RVI + LinReg)
Donald Dorsey's Inertia — a LinearRegression smoothing of the RVI
series. Endpoint of an n-bar least-squares fit of RVI is the indicator
reading. Preserves trend direction while damping the ratio. Candle
input, two parameters (rvi_period, linreg_period) with defaults 14 / 20.
Touchpoints: inertia.rs + mod.rs + lib.rs re-export, PyInertia +
__init__.py + test_new_indicators CANDLE_SCALAR + test_known_values
constant reference, InertiaNode (4-column OHLC batch) + index.d.ts /
index.js + indicators.test.js factory + reference, WasmInertia,
candle-fuzz target, README + CHANGELOG.
* test(kst): Move KST out of MULTI dict (it is scalar-input)
KST sits in the MULTI dict (candle-input, multi-output) but its
update() takes a single f64, not a candle tuple. The shared streaming
loop in test_multi_streaming_matches_batch fed the OHLCV tuple in,
which crashed with `TypeError: argument 'value': must be real number,
not tuple` on every Python matrix entry.
Split into a new MULTI_SCALAR_INPUT dict with its own test function
that feeds the close-price stream as floats. KST is currently the
only such indicator; structure is ready for future scalar-input
multi-output additions (e.g. some MACD-shaped indicators).
* test(coverage): Cover SMI zero-range and ConnorsRsi zero-prev cold paths
codecov/patch on PR 40 flagged two uncovered defensive branches:
- SMI returns self.current early when the smoothed range collapses to
zero (`r2 <= 0.0`) so the formula stays defined. Exercised by feeding
bars where high == low.
- ConnorsRsi skips the ROC ring-buffer update when the previous price
is exactly zero so the divide-by-zero in `(input - prev) / prev` is
impossible. Exercised by seeding the first bar at 0.0.
|
||
|
wickra-bot
|
a39adb9dae | chore: sync indicator count to 78 [skip ci] | ||
|
kingchenc
|
1cd5d1d8da |
fix(ci): Drop site/index.md from sync-about workflow (#55)
site/ is local-only (listed in .git/info/exclude), so the sed call in the Patch step aborted the workflow on every push to main with "sed: can't read site/index.md: No such file or directory". That kept the README counter from being committed and skipped the wiki sync. Patches README only now. site/ stays out of CI until the marketing site is promoted. |
||
|
kingchenc
|
466faddd87 |
feat: Family 01 Moving Averages — ALMA / McGinley / FRAMA / VIDYA / JMA / Alligator / EVWMA (#39)
* feat(alma): add Arnaud Legoux Moving Average
Gaussian-weighted moving average with configurable centre (offset in
[0, 1]) and kernel width (sigma > 0). Pre-computes normalised weights
at construction so each update is a single rolling window dot product.
Reference: Arnaud Legoux and Dimitrios Kouzis-Loukas, 2009.
Touchpoints:
- crates/wickra-core: alma.rs + mod.rs + lib.rs re-export
- bindings/python: PyAlma + __init__.py + test_new_indicators +
test_known_values reference
- bindings/node: AlmaNode + index.d.ts/index.js + indicators.test.js
factory + reference value
- bindings/wasm: wasm_scalar_indicator! macro
- fuzz: indicator_update target covers ALMA(9, 0.85, 6.0)
- crates/wickra/benches: bench_scalar entry
- README + CHANGELOG: Moving Averages row + Unreleased entry
* feat(mcginley): add McGinley Dynamic moving average
John McGinley's self-adjusting moving average with the recurrence
MD + (price - MD) / (0.6 * period * (price / MD)^4). Speeds up when
price falls below the indicator and damps when price runs above the
indicator. Seeded with the simple average of the first period inputs.
Reference: McGinley, Technical Analysis of Stocks & Commodities, 1990.
Touchpoints:
- crates/wickra-core: mcginley_dynamic.rs + mod.rs + lib.rs re-export
- bindings/python: PyMcGinleyDynamic + __init__.py + test_new_indicators
+ test_known_values reference
- bindings/node: McGinleyDynamicNode (scalar macro) + index.d.ts/index.js
+ indicators.test.js factory + reference value
- bindings/wasm: wasm_scalar_indicator! macro
- fuzz: indicator_update target covers McGinleyDynamic(10)
- crates/wickra/benches: bench_scalar entry
- README + CHANGELOG: Moving Averages row + Unreleased entry
* feat(frama): add Fractal Adaptive Moving Average
Ehlers' FRAMA adapts its smoothing constant to the fractal dimension of
the recent window: tight tracking in trends, heavy smoothing in chop.
Uses the close-only variant where max/min over each window half drive
the dimension estimate. Period must be even (default 16).
Reference: Ehlers, Fractal Adaptive Moving Average, 2005.
Touchpoints:
- crates/wickra-core: frama.rs + mod.rs + lib.rs re-export
- bindings/python: PyFrama + __init__.py + test_new_indicators +
test_known_values reference (constant series + uptrend tracking)
- bindings/node: FramaNode (scalar macro) + index.d.ts/index.js +
indicators.test.js factory + reference value
- bindings/wasm: wasm_scalar_indicator! macro
- fuzz: indicator_update target covers Frama(16)
- crates/wickra/benches: bench_scalar entry
- README + CHANGELOG: Moving Averages row + Unreleased entry
* feat(vidya): add Variable Index Dynamic Average
Chande's VIDYA — an EMA whose alpha scales with |CMO(cmo_period)| / 100.
Strong directional momentum lifts the smoothing constant toward the
EMA-of-period rate; flat or choppy windows shrink it toward zero so
VIDYA coasts on its previous value. Two parameters: period (14) and
cmo_period (9). Reuses the existing wickra-core Cmo internally.
Reference: Chande, Stocks & Commodities, 1992.
Also fixes a silent gap from
|
||
|
kingchenc
|
178fbfd68e |
ci: Add sync-about workflow to auto-update indicator count (#38)
Counts `mod xxx;` declarations in crates/wickra-core/src/indicators/mod.rs on every push to main, every PR, and every v* tag push. On non-PR runs it syncs the count into: - GitHub repo About description (via `gh repo edit`) - README.md + site/index.md (commit with [skip ci] back to main) - Wiki: Home.md, FAQ.md, Streaming-vs-Batch.md Requires a `ABOUT_SYNC_TOKEN` secret (classic PAT with `repo` scope, or fine-grained PAT with Administration+Contents write on the wickra repo). PR runs are read-only: count is logged but nothing is mutated, so forks cannot trigger writes. |
||
|
kingchenc
|
e30b3c6b35 |
release: 0.2.7 (Windows ARM64 restored + CPU label fix) (#37)
* chore(docs): rename benchmark CPU from 7950X3D to 9950X
The "Reproduced on" line in the umbrella + binding READMEs and the
benchmark page on the site listed the wrong AMD CPU. The benchmarks
were actually produced on a Ryzen 9 9950X, not a 7950X3D. Same
column for absolute µs values applies — the speedup ratios in the
tables are unchanged either way because they're relative across
libraries on the same machine.
The performance-regression issue template's CPU example also
updated for consistency (it was a generic placeholder, but matching
the canonical machine makes the example concrete).
* chore(npm): restore Windows ARM64 sub-package + napi matrix entry
npm Support unblocked the `wickra-win32-arm64-msvc` package name and
transferred write access to @kingchenc (placeholder 0.0.1-security
was published from their side; we ship our first real version on
top of that). This re-enables every change
v0.2.7
|
||
|
kingchenc
|
070be2eb27 |
release: 0.2.6 (docs.rs fix + README table reordering) (#36)
* fix(docs-rs): rename `doc_auto_cfg` to `doc_cfg` after Rust 1.92 merge `doc_auto_cfg` was removed in Rust 1.92.0 and folded back into `doc_cfg` (rust-lang/rust#138907). docs.rs builds with the latest nightly and sets `--cfg docsrs`, so the previous #![cfg_attr(docsrs, feature(doc_auto_cfg))] aborts compilation with E0557 on every published 0.2.x. GitHub CI never tripped this — stable rustc ignores the line because nothing sets the `docsrs` cfg there. Switch all three published library crates (`wickra`, `wickra-core`, `wickra-data`) to the merged-into `doc_cfg` gate. Same intent, same on-docs.rs output, builds again on nightly. * docs(readme): float Wickra to the top of the comparison tables Reorders the "Why Wickra exists" library-comparison table and the two benchmark headers so Wickra is the first row (with a ★ marker) instead of the last. The previous order placed Wickra at the bottom, which buries the only row a reader landing on the README is here to compare against. Same column data, same ★/winner annotations, just the row order flipped and a ★ prefix on the Wickra label. Mirrored across the umbrella README and every binding README so the crates.io / PyPI / npm landing pages stay in sync. * release: bump workspace + bindings to 0.2.6 Workspace, every binding (Python, Node, Node platform stubs), the release.yml comment and the CHANGELOG all move together to 0.2.6 so the next tagged release lines every artefact up. 0.2.6 carries two changes from the [0.2.6] CHANGELOG entry: - fix(docs-rs): swap the now-removed `doc_auto_cfg` feature gate for the merged-into `doc_cfg` so docs.rs nightly builds resume. - docs(readme): float ★ Wickra to the top of every comparison table across the umbrella + binding READMEs. wickra-win32-arm64-msvc stays excluded for this release with the same npm spam-filter rationale that held for 0.2.5.v0.2.6 |
||
|
kingchenc
|
221f7a71bc |
chore(github): add detailed issue & PR templates, capitalize title prefixes
Adds five new issue templates (bug_report_detailed, feature_request_detailed, performance_regression, documentation, question) alongside the existing short forms, plus an optional detailed PR template under .github/PULL_REQUEST_TEMPLATE/ that contributors can opt into via the ?template=detailed.md URL. Existing templates keep their behavior; only title prefixes and prose-paren wording were capitalized for consistency.v0.2.5 |
||
|
kingchenc
|
b5afc0a7e7 |
release: bump workspace + bindings to 0.2.5 (#35)
Workspace, every binding (Python, Node, Node platform stubs), and the release.yml comment are all updated together so the next tagged release on `v0.2.5` lines every artefact up. Also adds a short README "Disclaimer" section pointing out that Wickra is an indicator toolkit, not a trading system, and that production use is at the caller's own risk. The legal terms in LICENSE (PolyForm Noncommercial 1.0.0, "No Liability") already cover the warranty / as-is language — the README section just makes the trading-specific framing visible without burying it in a click-through. CHANGELOG carries the new 0.2.5 entry with the API addition (`BinanceConfig` + `connect_with_config`) and the best-effort Pong write change in `BinanceKlineStream::next_event`. wickra-win32-arm64-msvc stays excluded for this release with the same npm-spam-filter rationale that held for 0.2.1. |
||
|
kingchenc
|
9acb2f607e |
test(binance): mock-WS suite drives async/reconnect paths to ~100% (#34)
* refactor(binance): introduce BinanceConfig for endpoint + timing knobs Replaces the file-private READ_TIMEOUT / MAX_RECONNECT_ATTEMPTS / size limit constants with a Default-equipped BinanceConfig the caller can hand to a new connect_with_config(). connect() forwards to it with the defaults, so the public API stays backwards-compatible. Behaviour-preserving: every default matches the value of the constant it replaces, and the WebSocketConfig is built the same way. The change unlocks two real use-cases — pointing at Binance Testnet (wss://testnet.binance.vision) and pointing at a local mock server with millisecond-scale reconnect timing in tests. * test(binance): cover the Interval table and the empty-symbol guard Three quick wins that don't need a live or mock socket: - interval_as_str_covers_every_variant pins every wire-format mapping in one table so a typo on any of the 14 variants is caught. - binance_config_default_matches_production_endpoint guards the default base URL and timing knobs against an accidental drift. - connect_rejects_an_empty_symbol_list exercises the guard before the WebSocket handshake — the one async path we can hit without a server. * test(binance): cover the async / reconnect / control-frame paths Adds a small mock-WebSocket scaffold built on a `127.0.0.1:0` listener and tokio-tungstenite's `accept_async`, plus nine integration tests that drive `BinanceKlineStream::next_event` through every branch: - text + binary kline frames decode to a KlineEvent - inbound Ping is answered with a Pong, then the kline arrives - inbound Pong / Frame variants are silently skipped - a server-side Close triggers a transparent reconnect that then serves the kline - a stalled connection trips read_timeout and reconnects on its own - close() flips the closed flag and next_event() yields None forever - when every reconnect attempt is refused, next_event surfaces an Err - a "kline" envelope whose numbers are unparseable bubbles up as Error::Malformed rather than being silently skipped `one_shot_server` drops the listener as soon as the first accept is done, so a follow-up reconnect lands on a refused port — that is what lets the exhaustion test hit the final `last_err.expect(...)`. The whole suite runs in ~4 s with millisecond-scale reconnect timings supplied via the new test-only [`test_config`]. * test(binance): drop defensive cold-paths in the mock-WS scaffolding Codecov's patch report on PR #34 flagged seven uncovered lines, all of them in the test scaffolding rather than in production code: - the `let Ok((stream, _)) = … else { return }` shortcut and the `if let Ok(ws) = accept_async(stream).await { … }` branch in the mock-server helpers — both error arms never fire on a passing test - the closing braces of the spawned-task bodies in the close-frame and read-timeout reconnect tests — the spawned async blocks were getting killed mid-drain when the test asserted and returned Refactor the helpers to `.unwrap()` every Result (a failure here is a bug in the scaffold, not in production) and have `multi_shot_server` accept a fixed `n_accepts`, await every spawned inner task, and hand the outer JoinHandle back to the caller. Refactor the two affected tests to capture that JoinHandle, collapse the per-index `if/else` so both arms reach the same trailing expression, swap the read-timeout drain for a bounded sleep, and await `server_done` at the end. Every handler now reaches its closing brace before the runtime is torn down, so coverage on the patch should collapse from 97.89 % to 100 %. * test(binance): cover the non-kline-skip path and simplify the Ping arm After the scaffolding fix landed three lines on binance.rs were still uncovered: - L305 / L313: the Text- and Binary-arm "frame was not a kline, keep reading" fall-throughs. No existing test drove the loop through a non-kline frame followed by a kline; the new `next_event_skips_non_kline_frames_and_returns_the_next_kline` does exactly that (Text ack, Binary id frame, then a real kline). - L317: the Ping-Err defensive arm that forced a reconnect when the Pong reply itself failed to write. A failed Pong reply means the socket is already dead, so the very next read will surface the error and reconnect through the existing timeout/err branch — one tokio scheduling iteration later. Drop the defensive arm and write the Pong reply best-effort. Same observable behaviour, no test back door, no dead-line guard. Repo coverage on `cov/binance-mock-ws` now sits at 100 %. |
||
|
kingchenc
|
32caf023dd |
test(psar): drop violation-tuple cold path in trend tests (99.03 -> 100) (#33)
After PR #27 brought psar.rs to 99.03 %, Codecov still flagged the 'violation found' tuple arms in the trend tests (line 256 in pure_uptrend_sar_below_lows, line 285 in pure_downtrend_sar_above_highs) as missed: both tests are designed to NEVER find a violation, so the filter_map branch that constructs the (index, sar, bound) tuple is dead by design. Restructure both tests to use `.all(|(i, sar)| sar.is_none_or(|s| <bound>))` instead of collecting violations into a Vec. The closure runs on every emitted Some, asserts the SAR-vs-extreme bound directly, and the iterator short-circuits on the first false — no cold tuple construction left to count as uncovered. Semantics are identical (still asserts every SAR sits on the correct side of every candle's extreme); the diagnostic message loses the violating index list, which the tests never printed in any green run anyway. psar.rs is now at 207/207 lines, no behavioural change. |
||
|
kingchenc
|
250b75d468 |
test: 100% coverage for balance_of_power + median_price + true_range + typical_price + weighted_close (#32)
* test(balance_of_power): cover name metadata Codecov flagged 3 lines (file at 96.25%): Indicator-impl name body (73-75). * test(median_price): cover name metadata Codecov flagged 3 lines (file at 94.44%): Indicator-impl name body (62-64). * test(true_range): cover name metadata Codecov flagged 3 lines (file at 95.94%): Indicator-impl name body (73-75). * test(typical_price): cover name metadata Codecov flagged 3 lines (file at 94.44%): Indicator-impl name body (62-64). * test(weighted_close): cover name metadata Codecov flagged 3 lines (file at 94.44%): Indicator-impl name body (61-63). |
||
|
kingchenc
|
adc8488939 |
test: 100% coverage for ema + historical_volatility + kama + linreg_angle + mass_index (#24)
* test(ema): cover period accessor + warmup/name metadata Codecov flagged 9 lines in crates/wickra-core/src/indicators/ema.rs (file at 94.03%): const accessor period (74-77), Indicator-impl warmup_period (123-125), name (131-133). ema.rs now at 151/151. * test(historical_volatility): cover periods/value accessors + name metadata Codecov flagged 9 lines in crates/wickra-core/src/indicators/historical_volatility.rs (file at 93.87%): const accessors periods (80-83), value (85-88) and Indicator-impl name (153-155). historical_volatility.rs now at 147/147. * test(kama): cover periods accessor + warmup/name metadata Codecov flagged 9 lines in crates/wickra-core/src/indicators/kama.rs (file at 91.26%): accessor periods (65-67), Indicator-impl warmup_period (115-117), name (123-125). kama.rs now at 103/103. * test(linreg_angle): cover period accessor + warmup/name metadata Codecov flagged 9 lines in crates/wickra-core/src/indicators/linreg_angle.rs (file at 88.15%): const accessor period (50-52), Indicator-impl warmup_period (67-69), name (75-77). linreg_angle.rs now at 76/76. * test(mass_index): cover periods/value accessors + name metadata Codecov flagged 9 lines in crates/wickra-core/src/indicators/mass_index.rs (file at 91.42%): const accessors periods (80-82), value (85-87) and Indicator-impl name (134-136). mass_index.rs now at 105/105. |
||
|
kingchenc
|
d55d3db3d1 |
test: 100% coverage for vertical_horizontal_filter + z_score + vpt + csv + adl (#31)
* test(vertical_horizontal_filter): cover period accessor + name metadata Codecov flagged 6 lines (file at 94.44%): period (61-63) + name (119-121). * test(z_score): cover period accessor + name metadata Codecov flagged 6 lines (file at 93.75%): period (59-61) + name (106-108). * test(vpt): cover value() Some branch, name, zero-prev fallback Codecov flagged 5 lines (file at 94.38%): value() Some branch (57), prev==0.0 ROC fallback (77), and Indicator-impl name (100-102). Add accessors_and_metadata covering value()/name and zero_previous_ close_contributes_zero — feeding a 0.0 baseline + non-zero candle proves the divide-by-zero guard yields a 0 contribution rather than NaN. * test(csv): cover from_csv_reader + kill rejects_header dead panic arm Codecov flagged 5 lines in csv.rs (file at 96.98%): from_csv_reader (201-204) — never called by existing tests which use from_reader / open — and the cold arm in rejects_header_missing_a_column (279). Add from_csv_reader_accepts_a_ prebuilt_reader (demonstrates the API by building a custom-delimited csv::Reader and passing it in), and refactor the header-missing test to use a single matches!() assertion so the panic arm is gone. * test(adl): cover name metadata Codecov flagged 3 lines (file at 96.84%): Indicator-impl name body (94-96). |