millwright 0.2.1

A unified ML framework for Rust — proven Rust crates, assembled into one machine.
Documentation 
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
61
62
63
64
65
66
67
68
69
70
71
72
73
74
75
76
77
78
79
80
81
82
83
84
85
86
87
88
89
90
91
92
93
94
95
96
97
98
99
100
101
102
103
104
105
106
107
108
109
110
111
112
113
114
115
116
117
118
119
120
121
122
123
124
125
126
127
128
129
130
131
132
133
134
135
136
137
138
139
140
141
142
143
144
145
146
147
148
149
150
151
152
153
154
155
156

<html><head><meta http-equiv="Content-Type" content="text/html; charset=UTF-8"><title>Millwright · Docs</title>
<meta name="viewport" content="width=device-width, initial-scale=1">
<link rel="stylesheet" href="site.css">
</head><body><header class="top">
  <div class="wrap">
    <div class="brand"><a href="../index.html"><span class="mark">⚙</span>millwright</a><span class="ver">docs</span></div>
    <nav>
      <a href="index.html" class="active">home</a>
      <a href="data.html">data &amp; EDA</a>
      <a href="pipelines.html">pipelines</a>
      <a href="insight.html">insight</a>
      <a href="deploy.html">deploy</a>
      <a href="python.html">python</a>
      <a href="../index.html">design brief</a>
      <a class="repo" href="https://github.com/mi7plus/millwright">GitHub ↗</a>
    </nav>
  </div>
</header>

<main>
  <div class="wrap">
    <div class="hero">
      <div class="eyebrow">The guide</div>
      <h1>Data in, monitored<br>service out.</h1>
      <div class="sub">one data model · one contract · one pipeline</div>
      <p class="lede">The <a href="../index.html">design brief</a> is the <b>why</b>. These docs are the <b>how</b>: a hands-on run through the whole lifecycle, one topic per page. Every snippet is real API, mirrored from the runnable programs in <code class="inl">examples/</code>.</p>
    </div>
  </div>

  <!-- QUICKSTART -->
  <section id="quickstart">
    <div class="wrap">
      <div class="head col">
        <div class="eyebrow">Quickstart</div>
        <h2>Fit and predict in a dozen lines.</h2>
        <p class="muted">Add the crate — <code class="inl">default</code> is a lean, useful core (smartcore backend, preprocessing, model selection, ensembles).</p>
      </div>
<pre>cargo add millwright</pre>
<pre><span class="k">use</span> millwright::prelude::*;

<span class="c">// features as rows + a target -&gt; a Dataset</span>
<span class="k">let</span> x = <span class="f">Frame</span>::from_rows(rows, <span class="f">vec!</span>[<span class="s">"a"</span>.into(), <span class="s">"b"</span>.into()])?;
<span class="k">let</span> train = <span class="f">Dataset</span>::new(x, y)?;

<span class="c">// standardize, then a random forest — one composable object</span>
<span class="k">let mut</span> pipe = <span class="f">Pipeline</span>::new()
    .step(<span class="s">"scale"</span>, <span class="f">StandardScaler</span>::new())
    .estimator(<span class="s">"rf"</span>, <span class="f">RandomForest</span>::new());

pipe.fit(&amp;train)?;
<span class="k">let</span> preds = pipe.predict(&amp;test)?;</pre>
      <p class="run">cargo run --example spine</p>
    </div>
  </section>

  <!-- TOPIC CARDS -->
  <section id="topics">
    <div class="wrap">
      <div class="head col">
        <div class="eyebrow">The lifecycle, by topic</div>
        <h2>Where to go next.</h2>
      </div>
      <div class="cards">
        <a class="card" href="data.html">
          <div class="n">01 · data &amp; EDA</div>
          <h3>Frame, Dataset, Table &amp; Profile</h3>
          <p>The numeric boundary type, and the polars-backed typed layer that ingests CSV/Parquet, profiles it, and drafts a pipeline.</p>
        </a>
        <a class="card" href="pipelines.html">
          <div class="n">02 · pipelines &amp; models</div>
          <h3>The contract, preprocessing, search, ensembles</h3>
          <p>The four traits, composable pipelines tuned by path, cross-validation &amp; HPO, ensembles across backends, and a second backend (linfa).</p>
        </a>
        <a class="card" href="insight.html">
          <div class="n">03 · insight</div>
          <h3>Evaluate, explain, calibrate, detect</h3>
          <p>Metrics and diagnostics, SHAP, report figures, probability calibration, and unsupervised outlier detection.</p>
        </a>
        <a class="card" href="deploy.html">
          <div class="n">04 · deploy</div>
          <h3>ONNX, serving, registry, drift, AutoML</h3>
          <p>Export to one ONNX artifact, serve a drift-monitored endpoint, version models, and let AutoML search for the best deployable pipeline.</p>
        </a>
        <a class="card" href="python.html">
          <div class="n">05 · python</div>
          <h3><code class="inl">pip install millwright</code></h3>
          <p>The same Rust engine behind a Pythonic API, shipped as an abi3 wheel.</p>
        </a>
        <a class="card" href="https://docs.rs/millwright">
          <div class="n">reference ↗</div>
          <h3>API docs on docs.rs</h3>
          <p>Every type and method, generated from the source with all features.</p>
        </a>
      </div>
    </div>
  </section>

  <!-- FEATURES -->
  <section id="features">
    <div class="wrap">
      <div class="head col">
        <div class="eyebrow">Cargo features</div>
        <h2>Pull only what you need.</h2>
        <p class="muted">Every capability is a feature over one crate. A serving binary never compiles SHAP; a notebook never compiles <code class="inl">axum</code>. <code class="inl">full</code> lights up everything Rust-facing.</p>
      </div>
<pre><span class="c"># just the spine</span>
millwright = { version = <span class="s">"0.1"</span>, default-features = <span class="k">false</span>, features = [<span class="s">"smartcore-backend"</span>] }

<span class="c"># the whole lifecycle</span>
millwright = { version = <span class="s">"0.1"</span>, features = [<span class="s">"full"</span>] }</pre>
      <div class="tablewrap">
        <table>
          <thead><tr><th>Feature</th><th>Adds</th></tr></thead>
          <tbody>
            <tr class="core"><td><span class="feat">smartcore-backend</span><span class="badge">default</span></td><td>RandomForest · LinearRegression</td></tr>
            <tr class="core"><td><span class="feat">preprocessing</span><span class="badge">default</span></td><td>Smote · RandomOverSampler (imputers/scalers/encoders are core)</td></tr>
            <tr class="core"><td><span class="feat">model-selection</span><span class="badge">default</span></td><td>KFold · StratifiedKFold · GridSearch · RandomSearch · metrics</td></tr>
            <tr class="core"><td><span class="feat">ensemble</span><span class="badge">default</span></td><td>Voting · Bagging · Stacking</td></tr>
            <tr><td><span class="feat">eda</span></td><td>Table (polars CSV/Parquet ingest) · Profile (typed EDA)</td></tr>
            <tr><td><span class="feat">linfa-backend</span></td><td>KMeans · GaussianMixture · Dbscan · Pca</td></tr>
            <tr><td><span class="feat">hpo</span></td><td>BayesSearch (TPE) over a SearchSpace</td></tr>
            <tr><td><span class="feat">diagnostics</span></td><td>OLS Diagnostics: VIF · residuals · Cook's distance</td></tr>
            <tr><td><span class="feat">explain</span></td><td>Explainer (SHAP) · permutation_importance</td></tr>
            <tr><td><span class="feat">calibration</span></td><td>PlattScaling · IsotonicRegression · reliability_curve · CalibratedClassifier</td></tr>
            <tr><td><span class="feat">anomaly</span></td><td>Mahalanobis · KnnScore outlier detectors</td></tr>
            <tr><td><span class="feat">viz</span></td><td>ROC / residual SVG figures</td></tr>
            <tr><td><span class="feat">onnx</span></td><td>export_onnx · InferenceModel (tract)</td></tr>
            <tr><td><span class="feat">registry</span></td><td>versioned model Registry</td></tr>
            <tr><td><span class="feat">monitor</span></td><td>DriftMonitor (PSI)</td></tr>
            <tr><td><span class="feat">serve</span></td><td>Server — POST /predict, GET /metrics</td></tr>
            <tr><td><span class="feat">timeseries</span></td><td>AutoArima forecaster</td></tr>
            <tr><td><span class="feat">incremental</span></td><td>IncrementalLinear (partial_fit)</td></tr>
            <tr><td><span class="feat">automl</span></td><td>AutoML search</td></tr>
            <tr><td><span class="feat">python</span></td><td>the <code class="inl">pip install millwright</code> package</td></tr>
          </tbody>
        </table>
      </div>
      <p class="tiny">Reproducibility is a feature too: engines pinned to exact versions, a committed <code class="inl">Cargo.lock</code>, golden-output tests, and a feature-matrix CI. See the <a href="https://github.com/mi7plus/millwright">repo</a>.</p>
    </div>
  </section>

  <div class="wrap">
    <div class="pager">
      <span></span>
      <a class="next" href="data.html"><span class="dir">next →</span><b>Data &amp; EDA</b></a>
    </div>
  </div>
</main>

<footer>
  <div class="wrap">
    <span class="mono">⚙ millwright docs</span>
    <span class="mono"><a href="../index.html">design brief</a> · <a href="https://crates.io/crates/millwright">crates.io</a> · <a href="https://pypi.org/project/millwright/">PyPI</a> · <a href="https://docs.rs/millwright">docs.rs</a> · <a href="https://github.com/mi7plus/millwright">GitHub</a></span>
  </div>
</footer>
</body></html>