seldonio.github.com/python/seldon.pipeline.html at master-backup · SeldonIO/seldonio.github.com

924 lines (770 loc) · 60.2 KB
<!DOCTYPE html>
<!--[if IE 8]><html class="no-js lt-ie9" lang="en" > <![endif]-->
<!--[if gt IE 8]><!--> <html class="no-js" lang="en" > <!--<![endif]-->
  <meta charset="utf-8">
  <meta name="viewport" content="width=device-width, initial-scale=1.0">
  <title>seldon.pipeline package &mdash; seldon 2.2.1 documentation</title>
    <link rel="stylesheet" href="static/css/theme.css" type="text/css" />
    <link rel="top" title="seldon 2.2.1 documentation" href="index.html"/>
        <link rel="up" title="seldon package" href="seldon.html"/>
        <link rel="next" title="seldon.text package" href="seldon.text.html"/>
        <link rel="prev" title="seldon.misc package" href="seldon.misc.html"/> 
  <script src="static/js/modernizr.min.js"></script>
<body class="wy-body-for-nav" role="document">
  <div class="wy-grid-for-nav">
    <nav data-toggle="wy-nav-shift" class="wy-nav-side">
      <div class="wy-side-scroll">
        <div class="wy-side-nav-search">
            <a href="index.html" class="icon icon-home"> seldon
          </a>
              <div class="version">
              </div>
<div role="search">
  <form id="rtd-search-form" class="wy-form" action="search.html" method="get">
    <input type="text" name="q" placeholder="Search docs" />
    <input type="hidden" name="check_keywords" value="yes" />
    <input type="hidden" name="area" value="default" />
        </div>
        <div class="wy-menu wy-menu-vertical" data-spy="affix" role="navigation" aria-label="main navigation">
                <ul class="current">
<li class="toctree-l1 current"><a class="reference internal" href="seldon.html">seldon package</a><ul class="current">
<li class="toctree-l2 current"><a class="reference internal" href="seldon.html#subpackages">Subpackages</a><ul class="current">
<li class="toctree-l3"><a class="reference internal" href="seldon.anomaly.html">seldon.anomaly package</a></li>
<li class="toctree-l3"><a class="reference internal" href="seldon.cli.html">seldon.cli package</a></li>
<li class="toctree-l3"><a class="reference internal" href="seldon.luigi.html">seldon.luigi package</a></li>
<li class="toctree-l3"><a class="reference internal" href="seldon.microservice.html">seldon.microservice package</a></li>
<li class="toctree-l3"><a class="reference internal" href="seldon.misc.html">seldon.misc package</a></li>
<li class="toctree-l3 current"><a class="current reference internal" href="">seldon.pipeline package</a><ul>
<li class="toctree-l4"><a class="reference internal" href="#submodules">Submodules</a></li>
<li class="toctree-l4"><a class="reference internal" href="#module-seldon.pipeline.auto_transforms">seldon.pipeline.auto_transforms module</a></li>
<li class="toctree-l4"><a class="reference internal" href="#module-seldon.pipeline.basic_transforms">seldon.pipeline.basic_transforms module</a></li>
<li class="toctree-l4"><a class="reference internal" href="#seldon-pipeline-bayes-optimize-module">seldon.pipeline.bayes_optimize module</a></li>
<li class="toctree-l4"><a class="reference internal" href="#module-seldon.pipeline.cross_validation">seldon.pipeline.cross_validation module</a></li>
<li class="toctree-l4"><a class="reference internal" href="#module-seldon.pipeline.pandas_pipelines">seldon.pipeline.pandas_pipelines module</a></li>
<li class="toctree-l4"><a class="reference internal" href="#module-seldon.pipeline.sklearn_transform">seldon.pipeline.sklearn_transform module</a></li>
<li class="toctree-l4"><a class="reference internal" href="#module-seldon.pipeline.tfidf_transform">seldon.pipeline.tfidf_transform module</a></li>
<li class="toctree-l4"><a class="reference internal" href="#module-seldon.pipeline">Module contents</a></li>
<li class="toctree-l3"><a class="reference internal" href="seldon.text.html">seldon.text package</a></li>
<li class="toctree-l2"><a class="reference internal" href="seldon.html#submodules">Submodules</a></li>
<li class="toctree-l2"><a class="reference internal" href="seldon.html#module-seldon.fileutil">seldon.fileutil module</a></li>
<li class="toctree-l2"><a class="reference internal" href="seldon.html#module-seldon.keras">seldon.keras module</a></li>
<li class="toctree-l2"><a class="reference internal" href="seldon.html#module-seldon.sklearn_estimator">seldon.sklearn_estimator module</a></li>
<li class="toctree-l2"><a class="reference internal" href="seldon.html#module-seldon.util">seldon.util module</a></li>
<li class="toctree-l2"><a class="reference internal" href="seldon.html#module-seldon.vw">seldon.vw module</a></li>
<li class="toctree-l2"><a class="reference internal" href="seldon.html#module-seldon.xgb">seldon.xgb module</a></li>
<li class="toctree-l2"><a class="reference internal" href="seldon.html#module-seldon.tensorflow_wrapper">seldon.tensorflow_wrapper module</a></li>
<li class="toctree-l2"><a class="reference internal" href="seldon.html#module-seldon">Module contents</a></li>
        </div>
      </div>
    <section data-toggle="wy-nav-shift" class="wy-nav-content-wrap">
      <nav class="wy-nav-top" role="navigation" aria-label="top navigation">
        <i data-toggle="wy-nav-top" class="fa fa-bars"></i>
        <a href="index.html">seldon</a>
      </nav>
      <div class="wy-nav-content">
        <div class="rst-content">
<div role="navigation" aria-label="breadcrumbs navigation">
  <ul class="wy-breadcrumbs">
    <li><a href="index.html">Docs</a> &raquo;</li>
          <li><a href="seldon.html">seldon package</a> &raquo;</li>
    <li>seldon.pipeline package</li>
      <li class="wy-breadcrumbs-aside">
            <a href="sources/seldon.pipeline.txt" rel="nofollow"> View page source</a>
      </li>
          <div role="main" class="document" itemscope="itemscope" itemtype="http://schema.org/Article">
           <div itemprop="articleBody">
  <div class="section" id="seldon-pipeline-package">
<h1>seldon.pipeline package<a class="headerlink" href="#seldon-pipeline-package" title="Permalink to this headline">¶</a></h1>
<div class="section" id="submodules">
<h2>Submodules<a class="headerlink" href="#submodules" title="Permalink to this headline">¶</a></h2>
<div class="section" id="module-seldon.pipeline.auto_transforms">
<span id="seldon-pipeline-auto-transforms-module"></span><h2>seldon.pipeline.auto_transforms module<a class="headerlink" href="#module-seldon.pipeline.auto_transforms" title="Permalink to this headline">¶</a></h2>
<dl class="class">
<dt id="seldon.pipeline.auto_transforms.AutoTransform">
<em class="property">class </em><code class="descclassname">seldon.pipeline.auto_transforms.</code><code class="descname">AutoTransform</code><span class="sig-paren">(</span><em>exclude=[], include=None, max_values_numeric_categorical=0, date_cols=[], custom_date_formats=None, ignore_vals=None, force_categorical=[], min_cat_percent=0.0, max_cat_percent=1.0, bool_map={'1': 1, '0': 0, 'false': 0, 'no': 0, '1.0': 1, 'yes': 1, 'true': 1, '0.0': 0}, cat_missing_val='UKN', date_transforms=[True, True, True, True], create_date_differences=False, nan_threshold=None, drop_constant_features=True, drop_duplicate_cols=True, min_max_limit=False</em><span class="sig-paren">)</span><a class="reference internal" href="modules/seldon/pipeline/auto_transforms.html#AutoTransform"><span class="viewcode-link">[source]</span></a><a class="headerlink" href="#seldon.pipeline.auto_transforms.AutoTransform" title="Permalink to this definition">¶</a></dt>
<dd><p>Bases: <code class="xref py py-class docutils literal"><span class="pre">sklearn.base.BaseEstimator</span></code>, <code class="xref py py-class docutils literal"><span class="pre">sklearn.base.TransformerMixin</span></code></p>
<p>Automatically transform a set of features into normalzied numeric or categorical features or dates</p>
<table class="docutils field-list" frame="void" rules="none">
<col class="field-name" />
<col class="field-body" />
<tbody valign="top">
<tr class="field-odd field"><th class="field-name">Parameters:</th><td class="field-body"><ul class="first last simple">
<li><strong>exclude</strong> (<em>list str, optional</em>) &#8211; list of features to not include</li>
<li><strong>include</strong> (<em>list str, optional</em>) &#8211; features to include if None then all unless exclude used</li>
<li><strong>max_values_numeric_categorical</strong> (<em>int, optional</em>) &#8211; max number of unique values for numeric feature to treat as categorical</li>
<li><strong>custom_date_formats</strong> (<em>list str, optional</em>) &#8211; list of custom date formats to try</li>
<li><strong>ignore_vals</strong> (<em>list str, optional</em>) &#8211; list of feature values to treat as NA/ignored values</li>
<li><strong>force_categorical</strong> (<em>list str, optional</em>) &#8211; features to force to be categorical</li>
<li><strong>min_cat_percent</strong> (<em>list str, optional</em>) &#8211; min percentage for a categorical value to be kept</li>
<li><strong>max_cat_percent</strong> (<em>float, optional</em>) &#8211; max percentage for a categorical value to be kept</li>
<li><strong>bool_map</strong> (<em>dict, optional</em>) &#8211; set of string values to be treated as boolean</li>
<li><strong>cat_missing_value</strong> (<em>str, optional</em>) &#8211; string to use for missing categorical values</li>
<li><strong>date_transforms</strong> (<em>list bool, optional</em>) &#8211; which date transforms to apply [hour,month,day_of_week,year], default is all</li>
<li><strong>create_date_differences</strong> (<em>bool, optional</em>) &#8211; whether to create differences between all date variables</li>
<li><strong>nan_threshold</strong> (<em>float, optional</em>) &#8211; featurs to drop if too many nan, threshold is between 0-1 as percent</li>
<li><strong>drop_constant_features</strong> (<em>bool, optional</em>) &#8211; drop a column if its value is constant</li>
<li><strong>duplicate columns</strong> (<em>drop</em>) &#8211; drop duplicate columns</li>
<li><strong>min_max_limit</strong> (<em>bool, optional</em>) &#8211; limit numeric cols to min and max seen in fit</li>
<dl class="method">
<dt id="seldon.pipeline.auto_transforms.AutoTransform.fit">
<code class="descname">fit</code><span class="sig-paren">(</span><em>df</em><span class="sig-paren">)</span><a class="reference internal" href="modules/seldon/pipeline/auto_transforms.html#AutoTransform.fit"><span class="viewcode-link">[source]</span></a><a class="headerlink" href="#seldon.pipeline.auto_transforms.AutoTransform.fit" title="Permalink to this definition">¶</a></dt>
<dd><p>Fit models against an input pandas dataframe</p>
<table class="docutils field-list" frame="void" rules="none">
<col class="field-name" />
<col class="field-body" />
<tbody valign="top">
<tr class="field-odd field"><th class="field-name">Parameters:</th><td class="field-body"><strong>X</strong> (<em>pandas dataframe</em>) &#8211; </td>
<tr class="field-even field"><th class="field-name">Returns:</th><td class="field-body"><strong>self</strong></td>
<tr class="field-odd field"><th class="field-name">Return type:</th><td class="field-body">object</td>
<dl class="method">
<dt id="seldon.pipeline.auto_transforms.AutoTransform.transform">
<code class="descname">transform</code><span class="sig-paren">(</span><em>df</em><span class="sig-paren">)</span><a class="reference internal" href="modules/seldon/pipeline/auto_transforms.html#AutoTransform.transform"><span class="viewcode-link">[source]</span></a><a class="headerlink" href="#seldon.pipeline.auto_transforms.AutoTransform.transform" title="Permalink to this definition">¶</a></dt>
<dd><p>transform a datframe with fitted models</p>
<table class="docutils field-list" frame="void" rules="none">
<col class="field-name" />
<col class="field-body" />
<tbody valign="top">
<tr class="field-odd field"><th class="field-name">Parameters:</th><td class="field-body"><strong>X</strong> (<em>pandas dataframe</em>) &#8211; </td>
<tr class="field-even field"><th class="field-name">Returns:</th><td class="field-body"></td>
<tr class="field-odd field"><th class="field-name">Return type:</th><td class="field-body">Transformed pandas dataframe</td>
<div class="section" id="module-seldon.pipeline.basic_transforms">
<span id="seldon-pipeline-basic-transforms-module"></span><h2>seldon.pipeline.basic_transforms module<a class="headerlink" href="#module-seldon.pipeline.basic_transforms" title="Permalink to this headline">¶</a></h2>
<dl class="class">
<dt id="seldon.pipeline.basic_transforms.BinaryTransform">
<em class="property">class </em><code class="descclassname">seldon.pipeline.basic_transforms.</code><code class="descname">BinaryTransform</code><span class="sig-paren">(</span><em>input_feature=None</em>, <em>output_feature=None</em><span class="sig-paren">)</span><a class="reference internal" href="modules/seldon/pipeline/basic_transforms.html#BinaryTransform"><span class="viewcode-link">[source]</span></a><a class="headerlink" href="#seldon.pipeline.basic_transforms.BinaryTransform" title="Permalink to this definition">¶</a></dt>
<dd><p>Bases: <code class="xref py py-class docutils literal"><span class="pre">sklearn.base.BaseEstimator</span></code>, <code class="xref py py-class docutils literal"><span class="pre">sklearn.base.TransformerMixin</span></code></p>
<p>Create a binary feature based on existence of another feature</p>
<table class="docutils field-list" frame="void" rules="none">
<col class="field-name" />
<col class="field-body" />
<tbody valign="top">
<tr class="field-odd field"><th class="field-name">Parameters:</th><td class="field-body"><ul class="first last simple">
<li><strong>input_feature</strong> (<em>str</em>) &#8211; input feature to transform</li>
<li><strong>output_feature</strong> (<em>str</em>) &#8211; output feature to place transformation</li>
<dl class="method">
<dt id="seldon.pipeline.basic_transforms.BinaryTransform.fit">
<code class="descname">fit</code><span class="sig-paren">(</span><em>X</em><span class="sig-paren">)</span><a class="reference internal" href="modules/seldon/pipeline/basic_transforms.html#BinaryTransform.fit"><span class="viewcode-link">[source]</span></a><a class="headerlink" href="#seldon.pipeline.basic_transforms.BinaryTransform.fit" title="Permalink to this definition">¶</a></dt>
<dd><p>nothing to do in fit</p>
<dl class="method">
<dt id="seldon.pipeline.basic_transforms.BinaryTransform.transform">
<code class="descname">transform</code><span class="sig-paren">(</span><em>df</em><span class="sig-paren">)</span><a class="reference internal" href="modules/seldon/pipeline/basic_transforms.html#BinaryTransform.transform"><span class="viewcode-link">[source]</span></a><a class="headerlink" href="#seldon.pipeline.basic_transforms.BinaryTransform.transform" title="Permalink to this definition">¶</a></dt>
<dd></dd></dl>
<dl class="class">
<dt id="seldon.pipeline.basic_transforms.ExcludeFeaturesTransform">
<em class="property">class </em><code class="descclassname">seldon.pipeline.basic_transforms.</code><code class="descname">ExcludeFeaturesTransform</code><span class="sig-paren">(</span><em>excluded=[]</em><span class="sig-paren">)</span><a class="reference internal" href="modules/seldon/pipeline/basic_transforms.html#ExcludeFeaturesTransform"><span class="viewcode-link">[source]</span></a><a class="headerlink" href="#seldon.pipeline.basic_transforms.ExcludeFeaturesTransform" title="Permalink to this definition">¶</a></dt>
<dd><p>Bases: <code class="xref py py-class docutils literal"><span class="pre">sklearn.base.BaseEstimator</span></code>, <code class="xref py py-class docutils literal"><span class="pre">sklearn.base.TransformerMixin</span></code></p>
<p>Filter a dataset and exclude specided set of features</p>
<table class="docutils field-list" frame="void" rules="none">
<col class="field-name" />
<col class="field-body" />
<tbody valign="top">
<tr class="field-odd field"><th class="field-name">Parameters:</th><td class="field-body"><strong>excluded</strong> (<em>list str</em>) &#8211; list of features to be excluded</td>
<dl class="method">
<dt id="seldon.pipeline.basic_transforms.ExcludeFeaturesTransform.fit">
<code class="descname">fit</code><span class="sig-paren">(</span><em>X</em><span class="sig-paren">)</span><a class="reference internal" href="modules/seldon/pipeline/basic_transforms.html#ExcludeFeaturesTransform.fit"><span class="viewcode-link">[source]</span></a><a class="headerlink" href="#seldon.pipeline.basic_transforms.ExcludeFeaturesTransform.fit" title="Permalink to this definition">¶</a></dt>
<dd><p>nothing to do in fit</p>
<dl class="method">
<dt id="seldon.pipeline.basic_transforms.ExcludeFeaturesTransform.transform">
<code class="descname">transform</code><span class="sig-paren">(</span><em>df</em><span class="sig-paren">)</span><a class="reference internal" href="modules/seldon/pipeline/basic_transforms.html#ExcludeFeaturesTransform.transform"><span class="viewcode-link">[source]</span></a><a class="headerlink" href="#seldon.pipeline.basic_transforms.ExcludeFeaturesTransform.transform" title="Permalink to this definition">¶</a></dt>
<dd><p>Trasform dataframe to include specified features only</p>
<table class="docutils field-list" frame="void" rules="none">
<col class="field-name" />
<col class="field-body" />
<tbody valign="top">
<tr class="field-odd field"><th class="field-name">Parameters:</th><td class="field-body"><strong>df</strong> (<em>pandas dataframe</em>) &#8211; </td>
<tr class="field-even field"><th class="field-name">Returns:</th><td class="field-body"></td>
<tr class="field-odd field"><th class="field-name">Return type:</th><td class="field-body">Transformed pandas dataframe</td>
<dl class="class">
<dt id="seldon.pipeline.basic_transforms.ExistFeaturesTransform">
<em class="property">class </em><code class="descclassname">seldon.pipeline.basic_transforms.</code><code class="descname">ExistFeaturesTransform</code><span class="sig-paren">(</span><em>included=None</em><span class="sig-paren">)</span><a class="reference internal" href="modules/seldon/pipeline/basic_transforms.html#ExistFeaturesTransform"><span class="viewcode-link">[source]</span></a><a class="headerlink" href="#seldon.pipeline.basic_transforms.ExistFeaturesTransform" title="Permalink to this definition">¶</a></dt>
<dd><p>Bases: <code class="xref py py-class docutils literal"><span class="pre">sklearn.base.BaseEstimator</span></code>, <code class="xref py py-class docutils literal"><span class="pre">sklearn.base.TransformerMixin</span></code></p>
<p>Filter rows based on whether a specified set of features exists</p>
<table class="docutils field-list" frame="void" rules="none">
<col class="field-name" />
<col class="field-body" />
<tbody valign="top">
<tr class="field-odd field"><th class="field-name">Parameters:</th><td class="field-body"><strong>included</strong> (<em>list str</em>) &#8211; list of features that need to exist</td>
<dl class="method">
<dt id="seldon.pipeline.basic_transforms.ExistFeaturesTransform.fit">
<code class="descname">fit</code><span class="sig-paren">(</span><em>objs</em><span class="sig-paren">)</span><a class="reference internal" href="modules/seldon/pipeline/basic_transforms.html#ExistFeaturesTransform.fit"><span class="viewcode-link">[source]</span></a><a class="headerlink" href="#seldon.pipeline.basic_transforms.ExistFeaturesTransform.fit" title="Permalink to this definition">¶</a></dt>
<dd></dd></dl>
<dl class="method">
<dt id="seldon.pipeline.basic_transforms.ExistFeaturesTransform.transform">
<code class="descname">transform</code><span class="sig-paren">(</span><em>df</em><span class="sig-paren">)</span><a class="reference internal" href="modules/seldon/pipeline/basic_transforms.html#ExistFeaturesTransform.transform"><span class="viewcode-link">[source]</span></a><a class="headerlink" href="#seldon.pipeline.basic_transforms.ExistFeaturesTransform.transform" title="Permalink to this definition">¶</a></dt>
<dd><p>Transform by returning input feature set if required features exist in it</p>
<table class="docutils field-list" frame="void" rules="none">
<col class="field-name" />
<col class="field-body" />
<tbody valign="top">
<tr class="field-odd field"><th class="field-name">Parameters:</th><td class="field-body"><strong>df</strong> (<em>pandas dataframe</em>) &#8211; </td>
<tr class="field-even field"><th class="field-name">Returns:</th><td class="field-body"></td>
<tr class="field-odd field"><th class="field-name">Return type:</th><td class="field-body">Transformed pandas dataframe</td>
<dl class="class">
<dt id="seldon.pipeline.basic_transforms.FeatureIdTransform">
<em class="property">class </em><code class="descclassname">seldon.pipeline.basic_transforms.</code><code class="descname">FeatureIdTransform</code><span class="sig-paren">(</span><em>input_feature=None</em>, <em>output_feature=None</em>, <em>min_size=0</em>, <em>max_classes=1000</em>, <em>exclude_missing=False</em>, <em>zero_based=False</em>, <em>id_map={}</em><span class="sig-paren">)</span><a class="reference internal" href="modules/seldon/pipeline/basic_transforms.html#FeatureIdTransform"><span class="viewcode-link">[source]</span></a><a class="headerlink" href="#seldon.pipeline.basic_transforms.FeatureIdTransform" title="Permalink to this definition">¶</a></dt>
<dd><p>Bases: <code class="xref py py-class docutils literal"><span class="pre">sklearn.base.BaseEstimator</span></code>, <code class="xref py py-class docutils literal"><span class="pre">sklearn.base.TransformerMixin</span></code></p>
<p>create a numeric feature id</p>
<table class="docutils field-list" frame="void" rules="none">
<col class="field-name" />
<col class="field-body" />
<tbody valign="top">
<tr class="field-odd field"><th class="field-name">Parameters:</th><td class="field-body"><ul class="first last simple">
<li><strong>input_feature</strong> (<em>str</em>) &#8211; input feature to create ids from</li>
<li><strong>output_feature</strong> (<em>str</em>) &#8211; output feature to place ids</li>
<li><strong>min_size</strong> (<em>int, optional</em>) &#8211; minimum number of examples of each feature value for feature to be included in transform as new id</li>
<li><strong>exclude_missing</strong> (<em>bool, optional</em>) &#8211; exclude rows that do not have the input feature</li>
<dl class="method">
<dt id="seldon.pipeline.basic_transforms.FeatureIdTransform.fit">
<code class="descname">fit</code><span class="sig-paren">(</span><em>df</em><span class="sig-paren">)</span><a class="reference internal" href="modules/seldon/pipeline/basic_transforms.html#FeatureIdTransform.fit"><span class="viewcode-link">[source]</span></a><a class="headerlink" href="#seldon.pipeline.basic_transforms.FeatureIdTransform.fit" title="Permalink to this definition">¶</a></dt>
<dd><p>Create map of ids for each feature value</p>
<p>create counts of occurrences of each feature value. Exclude features with not enough counds. Create id map.</p>
<table class="docutils field-list" frame="void" rules="none">
<col class="field-name" />
<col class="field-body" />
<tbody valign="top">
<tr class="field-odd field"><th class="field-name">Parameters:</th><td class="field-body"><strong>df</strong> (<em>pandas dataframe</em>) &#8211; </td>
<tr class="field-even field"><th class="field-name">Returns:</th><td class="field-body"><strong>self</strong></td>
<tr class="field-odd field"><th class="field-name">Return type:</th><td class="field-body">object</td>
<dl class="method">
<dt id="seldon.pipeline.basic_transforms.FeatureIdTransform.transform">
<code class="descname">transform</code><span class="sig-paren">(</span><em>df</em><span class="sig-paren">)</span><a class="reference internal" href="modules/seldon/pipeline/basic_transforms.html#FeatureIdTransform.transform"><span class="viewcode-link">[source]</span></a><a class="headerlink" href="#seldon.pipeline.basic_transforms.FeatureIdTransform.transform" title="Permalink to this definition">¶</a></dt>
<dd><p>Transform features creating a new id and exluding rows if needed</p>
<table class="docutils field-list" frame="void" rules="none">
<col class="field-name" />
<col class="field-body" />
<tbody valign="top">
<tr class="field-odd field"><th class="field-name">Parameters:</th><td class="field-body"><strong>df</strong> (<em>pandas dataframe</em>) &#8211; </td>
<tr class="field-even field"><th class="field-name">Returns:</th><td class="field-body"></td>
<tr class="field-odd field"><th class="field-name">Return type:</th><td class="field-body">Transformed pandas dataframe</td>
<dl class="class">
<dt id="seldon.pipeline.basic_transforms.IncludeFeaturesTransform">
<em class="property">class </em><code class="descclassname">seldon.pipeline.basic_transforms.</code><code class="descname">IncludeFeaturesTransform</code><span class="sig-paren">(</span><em>included=[]</em><span class="sig-paren">)</span><a class="reference internal" href="modules/seldon/pipeline/basic_transforms.html#IncludeFeaturesTransform"><span class="viewcode-link">[source]</span></a><a class="headerlink" href="#seldon.pipeline.basic_transforms.IncludeFeaturesTransform" title="Permalink to this definition">¶</a></dt>
<dd><p>Bases: <code class="xref py py-class docutils literal"><span class="pre">sklearn.base.BaseEstimator</span></code>, <code class="xref py py-class docutils literal"><span class="pre">sklearn.base.TransformerMixin</span></code></p>
<p>Filter a dataset and include only specided set of features</p>
<table class="docutils field-list" frame="void" rules="none">
<col class="field-name" />
<col class="field-body" />
<tbody valign="top">
<tr class="field-odd field"><th class="field-name">Parameters:</th><td class="field-body"><strong>input_features</strong> (<em>list str</em>) &#8211; input features to include</td>
<dl class="method">
<dt id="seldon.pipeline.basic_transforms.IncludeFeaturesTransform.fit">
<code class="descname">fit</code><span class="sig-paren">(</span><em>X</em><span class="sig-paren">)</span><a class="reference internal" href="modules/seldon/pipeline/basic_transforms.html#IncludeFeaturesTransform.fit"><span class="viewcode-link">[source]</span></a><a class="headerlink" href="#seldon.pipeline.basic_transforms.IncludeFeaturesTransform.fit" title="Permalink to this definition">¶</a></dt>
<dd><p>nothing to do in fit</p>
<dl class="method">
<dt id="seldon.pipeline.basic_transforms.IncludeFeaturesTransform.transform">
<code class="descname">transform</code><span class="sig-paren">(</span><em>df</em><span class="sig-paren">)</span><a class="reference internal" href="modules/seldon/pipeline/basic_transforms.html#IncludeFeaturesTransform.transform"><span class="viewcode-link">[source]</span></a><a class="headerlink" href="#seldon.pipeline.basic_transforms.IncludeFeaturesTransform.transform" title="Permalink to this definition">¶</a></dt>
<dd><p>transform a dataframe to include given features</p>
<table class="docutils field-list" frame="void" rules="none">
<col class="field-name" />
<col class="field-body" />
<tbody valign="top">
<tr class="field-odd field"><th class="field-name">Parameters:</th><td class="field-body"><strong>df</strong> (<em>pandas dataframe</em>) &#8211; </td>
<tr class="field-even field"><th class="field-name">Returns:</th><td class="field-body"></td>
<tr class="field-odd field"><th class="field-name">Return type:</th><td class="field-body">Transformed pandas dataframe</td>
<dl class="class">
<dt id="seldon.pipeline.basic_transforms.SplitTransform">
<em class="property">class </em><code class="descclassname">seldon.pipeline.basic_transforms.</code><code class="descname">SplitTransform</code><span class="sig-paren">(</span><em>split_expression=' '</em>, <em>ignore_numbers=False</em>, <em>input_features=[]</em>, <em>output_feature=None</em><span class="sig-paren">)</span><a class="reference internal" href="modules/seldon/pipeline/basic_transforms.html#SplitTransform"><span class="viewcode-link">[source]</span></a><a class="headerlink" href="#seldon.pipeline.basic_transforms.SplitTransform" title="Permalink to this definition">¶</a></dt>
<dd><p>Bases: <code class="xref py py-class docutils literal"><span class="pre">sklearn.base.BaseEstimator</span></code>, <code class="xref py py-class docutils literal"><span class="pre">sklearn.base.TransformerMixin</span></code></p>
<p>Split a set of string input features on an expression and create a new feature which has a list of values</p>
<table class="docutils field-list" frame="void" rules="none">
<col class="field-name" />
<col class="field-body" />
<tbody valign="top">
<tr class="field-odd field"><th class="field-name">Parameters:</th><td class="field-body"><ul class="first last simple">
<li><strong>split_expression</strong> (<em>str</em>) &#8211; regular expression to split feature on</li>
<li><strong>ignore_numbers</strong> (<em>bool</em>) &#8211; whether to ignore any resulting strings that represent numbers</li>
<li><strong>input_features</strong> (<em>list str</em>) &#8211; list of feature names to split - should all have text values</li>
<li><strong>output_feature</strong> (<em>str</em>) &#8211; output feature</li>
<dl class="method">
<dt id="seldon.pipeline.basic_transforms.SplitTransform.fit">
<code class="descname">fit</code><span class="sig-paren">(</span><em>X</em><span class="sig-paren">)</span><a class="reference internal" href="modules/seldon/pipeline/basic_transforms.html#SplitTransform.fit"><span class="viewcode-link">[source]</span></a><a class="headerlink" href="#seldon.pipeline.basic_transforms.SplitTransform.fit" title="Permalink to this definition">¶</a></dt>
<dd></dd></dl>
<dl class="method">
<dt id="seldon.pipeline.basic_transforms.SplitTransform.transform">
<code class="descname">transform</code><span class="sig-paren">(</span><em>df</em><span class="sig-paren">)</span><a class="reference internal" href="modules/seldon/pipeline/basic_transforms.html#SplitTransform.transform"><span class="viewcode-link">[source]</span></a><a class="headerlink" href="#seldon.pipeline.basic_transforms.SplitTransform.transform" title="Permalink to this definition">¶</a></dt>
<dd><p>Transform text features by splitting them and creating a list of feature as result</p>
<table class="docutils field-list" frame="void" rules="none">
<col class="field-name" />
<col class="field-body" />
<tbody valign="top">
<tr class="field-odd field"><th class="field-name">Parameters:</th><td class="field-body"><strong>df</strong> (<em>pandas dataframe</em>) &#8211; </td>
<tr class="field-even field"><th class="field-name">Returns:</th><td class="field-body"></td>
<tr class="field-odd field"><th class="field-name">Return type:</th><td class="field-body">Transformed pandas dataframe</td>
<dl class="class">
<dt id="seldon.pipeline.basic_transforms.SvmlightTransform">
<em class="property">class </em><code class="descclassname">seldon.pipeline.basic_transforms.</code><code class="descname">SvmlightTransform</code><span class="sig-paren">(</span><em>included=None</em>, <em>zero_based=False</em>, <em>excluded=[]</em>, <em>id_map={}</em>, <em>output_feature=None</em>, <em>id_map_file=None</em><span class="sig-paren">)</span><a class="reference internal" href="modules/seldon/pipeline/basic_transforms.html#SvmlightTransform"><span class="viewcode-link">[source]</span></a><a class="headerlink" href="#seldon.pipeline.basic_transforms.SvmlightTransform" title="Permalink to this definition">¶</a></dt>
<dd><p>Bases: <code class="xref py py-class docutils literal"><span class="pre">sklearn.base.BaseEstimator</span></code>, <code class="xref py py-class docutils literal"><span class="pre">sklearn.base.TransformerMixin</span></code></p>
<p>Take a set of features and transform into a sorted dictionary of numeric id:value features</p>
<table class="docutils field-list" frame="void" rules="none">
<col class="field-name" />
<col class="field-body" />
<tbody valign="top">
<tr class="field-odd field"><th class="field-name">Parameters:</th><td class="field-body"><ul class="first last simple">
<li><strong>included</strong> (<em>list str</em>) &#8211; set of feature to use as input</li>
<li><strong>zero_based</strong> (<em>zero_based, optional</em>) &#8211; whether to start first id at 0</li>
<li><strong>excluded</strong> (<em>list str</em>) &#8211; set of features to exclude</li>
<dl class="method">
<dt id="seldon.pipeline.basic_transforms.SvmlightTransform.fit">
<code class="descname">fit</code><span class="sig-paren">(</span><em>df</em><span class="sig-paren">)</span><a class="reference internal" href="modules/seldon/pipeline/basic_transforms.html#SvmlightTransform.fit"><span class="viewcode-link">[source]</span></a><a class="headerlink" href="#seldon.pipeline.basic_transforms.SvmlightTransform.fit" title="Permalink to this definition">¶</a></dt>
<dd><p>create ids for each feature to be included</p>
<table class="docutils field-list" frame="void" rules="none">
<col class="field-name" />
<col class="field-body" />
<tbody valign="top">
<tr class="field-odd field"><th class="field-name">Parameters:</th><td class="field-body"><strong>df</strong> (<em>pandas dataframe</em>) &#8211; </td>
<tr class="field-even field"><th class="field-name">Returns:</th><td class="field-body"><strong>self</strong></td>
<tr class="field-odd field"><th class="field-name">Return type:</th><td class="field-body">object</td>
<dl class="method">
<dt id="seldon.pipeline.basic_transforms.SvmlightTransform.transform">
<code class="descname">transform</code><span class="sig-paren">(</span><em>df</em><span class="sig-paren">)</span><a class="reference internal" href="modules/seldon/pipeline/basic_transforms.html#SvmlightTransform.transform"><span class="viewcode-link">[source]</span></a><a class="headerlink" href="#seldon.pipeline.basic_transforms.SvmlightTransform.transform" title="Permalink to this definition">¶</a></dt>
<dd><p>Transform features by getting id and numeric value</p>
<table class="docutils field-list" frame="void" rules="none">
<col class="field-name" />
<col class="field-body" />
<tbody valign="top">
<tr class="field-odd field"><th class="field-name">Parameters:</th><td class="field-body"><strong>X</strong> (<em>pandas dataframe</em>) &#8211; </td>
<tr class="field-even field"><th class="field-name">Returns:</th><td class="field-body"></td>
<tr class="field-odd field"><th class="field-name">Return type:</th><td class="field-body">Transformed pandas dataframe</td>
<div class="section" id="seldon-pipeline-bayes-optimize-module">
<h2>seldon.pipeline.bayes_optimize module<a class="headerlink" href="#seldon-pipeline-bayes-optimize-module" title="Permalink to this headline">¶</a></h2>
<div class="section" id="module-seldon.pipeline.cross_validation">
<span id="seldon-pipeline-cross-validation-module"></span><h2>seldon.pipeline.cross_validation module<a class="headerlink" href="#module-seldon.pipeline.cross_validation" title="Permalink to this headline">¶</a></h2>
<dl class="class">
<dt id="seldon.pipeline.cross_validation.SeldonKFold">
<em class="property">class </em><code class="descclassname">seldon.pipeline.cross_validation.</code><code class="descname">SeldonKFold</code><span class="sig-paren">(</span><em>clf=None</em>, <em>k=5</em>, <em>save_folds_folder=None</em>, <em>metric='accuracy'</em>, <em>random_state=1</em><span class="sig-paren">)</span><a class="reference internal" href="modules/seldon/pipeline/cross_validation.html#SeldonKFold"><span class="viewcode-link">[source]</span></a><a class="headerlink" href="#seldon.pipeline.cross_validation.SeldonKFold" title="Permalink to this definition">¶</a></dt>
<dd><p>Bases: <code class="xref py py-class docutils literal"><span class="pre">sklearn.base.BaseEstimator</span></code></p>
<p>Simple wrapper to provide cross validation test using estimator with input from pandas dataframe</p>
<table class="docutils field-list" frame="void" rules="none">
<col class="field-name" />
<col class="field-body" />
<tbody valign="top">
<tr class="field-odd field"><th class="field-name">Parameters:</th><td class="field-body"><ul class="first last simple">
<li><strong>clf</strong> (<em>object</em>) &#8211; Pandas compatible scikit learn Estimator to apply to data splits</li>
<li><strong>k</strong> (<em>int, optional</em>) &#8211; number of folds</li>
<li><strong>save_folder_folder</strong> (<em>str, optional</em>) &#8211; a folder to save prediction results from each fold</li>
<dl class="method">
<dt id="seldon.pipeline.cross_validation.SeldonKFold.fit">
<code class="descname">fit</code><span class="sig-paren">(</span><em>X</em>, <em>y=None</em><span class="sig-paren">)</span><a class="reference internal" href="modules/seldon/pipeline/cross_validation.html#SeldonKFold.fit"><span class="viewcode-link">[source]</span></a><a class="headerlink" href="#seldon.pipeline.cross_validation.SeldonKFold.fit" title="Permalink to this definition">¶</a></dt>
<dd><p>Split dataframe into k folds and train test classifier on each. Finally train classifier on all data.
:type X: pandas dataframe</p>
<table class="docutils field-list" frame="void" rules="none">
<col class="field-name" />
<col class="field-body" />
<tbody valign="top">
<tr class="field-odd field"><th class="field-name">Returns:</th><td class="field-body"><strong>self</strong></td>
<tr class="field-even field"><th class="field-name">Return type:</th><td class="field-body">object</td>
<dl class="method">
<dt id="seldon.pipeline.cross_validation.SeldonKFold.get_class_id_map">
<code class="descname">get_class_id_map</code><span class="sig-paren">(</span><span class="sig-paren">)</span><a class="reference internal" href="modules/seldon/pipeline/cross_validation.html#SeldonKFold.get_class_id_map"><span class="viewcode-link">[source]</span></a><a class="headerlink" href="#seldon.pipeline.cross_validation.SeldonKFold.get_class_id_map" title="Permalink to this definition">¶</a></dt>
<dd></dd></dl>
<dl class="method">
<dt id="seldon.pipeline.cross_validation.SeldonKFold.get_score">
<code class="descname">get_score</code><span class="sig-paren">(</span><span class="sig-paren">)</span><a class="reference internal" href="modules/seldon/pipeline/cross_validation.html#SeldonKFold.get_score"><span class="viewcode-link">[source]</span></a><a class="headerlink" href="#seldon.pipeline.cross_validation.SeldonKFold.get_score" title="Permalink to this definition">¶</a></dt>
<dd></dd></dl>
<dl class="method">
<dt id="seldon.pipeline.cross_validation.SeldonKFold.get_scores">
<code class="descname">get_scores</code><span class="sig-paren">(</span><span class="sig-paren">)</span><a class="reference internal" href="modules/seldon/pipeline/cross_validation.html#SeldonKFold.get_scores"><span class="viewcode-link">[source]</span></a><a class="headerlink" href="#seldon.pipeline.cross_validation.SeldonKFold.get_scores" title="Permalink to this definition">¶</a></dt>
<dd></dd></dl>
<dl class="method">
<dt id="seldon.pipeline.cross_validation.SeldonKFold.predict">
<code class="descname">predict</code><span class="sig-paren">(</span><em>X</em><span class="sig-paren">)</span><a class="reference internal" href="modules/seldon/pipeline/cross_validation.html#SeldonKFold.predict"><span class="viewcode-link">[source]</span></a><a class="headerlink" href="#seldon.pipeline.cross_validation.SeldonKFold.predict" title="Permalink to this definition">¶</a></dt>
<dd></dd></dl>
<dl class="method">
<dt id="seldon.pipeline.cross_validation.SeldonKFold.predict_proba">
<code class="descname">predict_proba</code><span class="sig-paren">(</span><em>X</em><span class="sig-paren">)</span><a class="reference internal" href="modules/seldon/pipeline/cross_validation.html#SeldonKFold.predict_proba"><span class="viewcode-link">[source]</span></a><a class="headerlink" href="#seldon.pipeline.cross_validation.SeldonKFold.predict_proba" title="Permalink to this definition">¶</a></dt>
<dd></dd></dl>
<dl class="method">
<dt id="seldon.pipeline.cross_validation.SeldonKFold.set_params">
<code class="descname">set_params</code><span class="sig-paren">(</span><em>**params</em><span class="sig-paren">)</span><a class="reference internal" href="modules/seldon/pipeline/cross_validation.html#SeldonKFold.set_params"><span class="viewcode-link">[source]</span></a><a class="headerlink" href="#seldon.pipeline.cross_validation.SeldonKFold.set_params" title="Permalink to this definition">¶</a></dt>
<dd></dd></dl>
<dl class="method">
<dt id="seldon.pipeline.cross_validation.SeldonKFold.transform">
<code class="descname">transform</code><span class="sig-paren">(</span><em>X</em><span class="sig-paren">)</span><a class="reference internal" href="modules/seldon/pipeline/cross_validation.html#SeldonKFold.transform"><span class="viewcode-link">[source]</span></a><a class="headerlink" href="#seldon.pipeline.cross_validation.SeldonKFold.transform" title="Permalink to this definition">¶</a></dt>
<dd><p>Do nothing and pass input back</p>
<div class="section" id="module-seldon.pipeline.pandas_pipelines">
<span id="seldon-pipeline-pandas-pipelines-module"></span><h2>seldon.pipeline.pandas_pipelines module<a class="headerlink" href="#module-seldon.pipeline.pandas_pipelines" title="Permalink to this headline">¶</a></h2>
<dl class="class">
<dt id="seldon.pipeline.pandas_pipelines.BasePandasEstimator">
<em class="property">class </em><code class="descclassname">seldon.pipeline.pandas_pipelines.</code><code class="descname">BasePandasEstimator</code><span class="sig-paren">(</span><em>target=None</em>, <em>target_readable=None</em>, <em>included=None</em>, <em>excluded=None</em>, <em>id_map={}</em><span class="sig-paren">)</span><a class="reference internal" href="modules/seldon/pipeline/pandas_pipelines.html#BasePandasEstimator"><span class="viewcode-link">[source]</span></a><a class="headerlink" href="#seldon.pipeline.pandas_pipelines.BasePandasEstimator" title="Permalink to this definition">¶</a></dt>
<dd><p>Bases: <code class="xref py py-class docutils literal"><span class="pre">object</span></code></p>
<p>Tools to help with Pandas based estimators.</p>
<table class="docutils field-list" frame="void" rules="none">
<col class="field-name" />
<col class="field-body" />
<tbody valign="top">
<tr class="field-odd field"><th class="field-name">Parameters:</th><td class="field-body"><ul class="first last simple">
<li><strong>target</strong> (<em>str</em>) &#8211; Target column</li>
<li><strong>target_readable</strong> (<em>str</em>) &#8211; More descriptive version of target variable</li>
<li><strong>included</strong> (<em>list str, optional</em>) &#8211; columns to include</li>
<li><strong>excluded</strong> (<em>list str, optional</em>) &#8211; columns to exclude</li>
<li><strong>id_map</strong> (<em>dict (int,str), optional</em>) &#8211; map of class ids to high level names</li>
<dl class="method">
<dt id="seldon.pipeline.pandas_pipelines.BasePandasEstimator.close">
<code class="descname">close</code><span class="sig-paren">(</span><span class="sig-paren">)</span><a class="reference internal" href="modules/seldon/pipeline/pandas_pipelines.html#BasePandasEstimator.close"><span class="viewcode-link">[source]</span></a><a class="headerlink" href="#seldon.pipeline.pandas_pipelines.BasePandasEstimator.close" title="Permalink to this definition">¶</a></dt>
<dd></dd></dl>
<dl class="method">
<dt id="seldon.pipeline.pandas_pipelines.BasePandasEstimator.convert_dataframe">
<code class="descname">convert_dataframe</code><span class="sig-paren">(</span><em>df_base</em>, <em>vectorizer</em><span class="sig-paren">)</span><a class="reference internal" href="modules/seldon/pipeline/pandas_pipelines.html#BasePandasEstimator.convert_dataframe"><span class="viewcode-link">[source]</span></a><a class="headerlink" href="#seldon.pipeline.pandas_pipelines.BasePandasEstimator.convert_dataframe" title="Permalink to this definition">¶</a></dt>
<dd><p>Convert a dataframe into one for use with ml algorithms
One hot encode the categorical variable
Ignore date values
concatenate with numeric values</p>
<dl class="method">
<dt id="seldon.pipeline.pandas_pipelines.BasePandasEstimator.convert_numpy">
<code class="descname">convert_numpy</code><span class="sig-paren">(</span><em>df</em><span class="sig-paren">)</span><a class="reference internal" href="modules/seldon/pipeline/pandas_pipelines.html#BasePandasEstimator.convert_numpy"><span class="viewcode-link">[source]</span></a><a class="headerlink" href="#seldon.pipeline.pandas_pipelines.BasePandasEstimator.convert_numpy" title="Permalink to this definition">¶</a></dt>
<dd><p>Convert a dataframe into a numpy data matrix for training and an array of target values
Uses a vectorizer for one hot encoding which is returned
NaNs are filled with zeros</p>
<table class="docutils field-list" frame="void" rules="none">
<col class="field-name" />
<col class="field-body" />
<tbody valign="top">
<tr class="field-odd field"><th class="field-name">Parameters:</th><td class="field-body"><strong>df</strong> (<em>pandas dataframe</em>) &#8211; </td>
<tr class="field-even field"><th class="field-name">Returns:</th><td class="field-body"><ul class="simple">
<li><strong>X</strong> (<em>array like - data as array</em>)</li>
<li><strong>y</strong> (<em>array - target labels</em>)</li>
<li><strong>vectorizer</strong> (<em>vectorizer used for one hot encoding</em>)</li>
<dl class="method">
<dt id="seldon.pipeline.pandas_pipelines.BasePandasEstimator.create_class_id_map">
<code class="descname">create_class_id_map</code><span class="sig-paren">(</span><em>df</em>, <em>target</em>, <em>target_readable</em>, <em>zero_based=True</em><span class="sig-paren">)</span><a class="reference internal" href="modules/seldon/pipeline/pandas_pipelines.html#BasePandasEstimator.create_class_id_map"><span class="viewcode-link">[source]</span></a><a class="headerlink" href="#seldon.pipeline.pandas_pipelines.BasePandasEstimator.create_class_id_map" title="Permalink to this definition">¶</a></dt>
<dd><p>Create a map of classification ids to readable values</p>
<dl class="method">
<dt id="seldon.pipeline.pandas_pipelines.BasePandasEstimator.encode_onehot">
<code class="descname">encode_onehot</code><span class="sig-paren">(</span><em>df</em>, <em>cols</em>, <em>vec</em>, <em>op</em><span class="sig-paren">)</span><a class="reference internal" href="modules/seldon/pipeline/pandas_pipelines.html#BasePandasEstimator.encode_onehot"><span class="viewcode-link">[source]</span></a><a class="headerlink" href="#seldon.pipeline.pandas_pipelines.BasePandasEstimator.encode_onehot" title="Permalink to this definition">¶</a></dt>
<dd><p>One hot encode categorical values from a data frame using a vectorizer passed in</p>
<dl class="method">
<dt id="seldon.pipeline.pandas_pipelines.BasePandasEstimator.get_class_id_map">
<code class="descname">get_class_id_map</code><span class="sig-paren">(</span><span class="sig-paren">)</span><a class="reference internal" href="modules/seldon/pipeline/pandas_pipelines.html#BasePandasEstimator.get_class_id_map"><span class="viewcode-link">[source]</span></a><a class="headerlink" href="#seldon.pipeline.pandas_pipelines.BasePandasEstimator.get_class_id_map" title="Permalink to this definition">¶</a></dt>
<dd></dd></dl>
<dl class="method">
<dt id="seldon.pipeline.pandas_pipelines.BasePandasEstimator.get_target">
<code class="descname">get_target</code><span class="sig-paren">(</span><span class="sig-paren">)</span><a class="reference internal" href="modules/seldon/pipeline/pandas_pipelines.html#BasePandasEstimator.get_target"><span class="viewcode-link">[source]</span></a><a class="headerlink" href="#seldon.pipeline.pandas_pipelines.BasePandasEstimator.get_target" title="Permalink to this definition">¶</a></dt>
<dd></dd></dl>
<dl class="method">
<dt id="seldon.pipeline.pandas_pipelines.BasePandasEstimator.predict">
<code class="descname">predict</code><span class="sig-paren">(</span><em>X</em><span class="sig-paren">)</span><a class="reference internal" href="modules/seldon/pipeline/pandas_pipelines.html#BasePandasEstimator.predict"><span class="viewcode-link">[source]</span></a><a class="headerlink" href="#seldon.pipeline.pandas_pipelines.BasePandasEstimator.predict" title="Permalink to this definition">¶</a></dt>
<dd></dd></dl>
<dl class="method">
<dt id="seldon.pipeline.pandas_pipelines.BasePandasEstimator.set_class_id_map">
<code class="descname">set_class_id_map</code><span class="sig-paren">(</span><em>id_map</em><span class="sig-paren">)</span><a class="reference internal" href="modules/seldon/pipeline/pandas_pipelines.html#BasePandasEstimator.set_class_id_map"><span class="viewcode-link">[source]</span></a><a class="headerlink" href="#seldon.pipeline.pandas_pipelines.BasePandasEstimator.set_class_id_map" title="Permalink to this definition">¶</a></dt>
<dd></dd></dl>
<div class="section" id="module-seldon.pipeline.sklearn_transform">
<span id="seldon-pipeline-sklearn-transform-module"></span><h2>seldon.pipeline.sklearn_transform module<a class="headerlink" href="#module-seldon.pipeline.sklearn_transform" title="Permalink to this headline">¶</a></h2>
<dl class="class">
<dt id="seldon.pipeline.sklearn_transform.SklearnTransform">
<em class="property">class </em><code class="descclassname">seldon.pipeline.sklearn_transform.</code><code class="descname">SklearnTransform</code><span class="sig-paren">(</span><em>input_features=None</em>, <em>output_features=None</em>, <em>output_features_prefix=None</em>, <em>transformer=None</em><span class="sig-paren">)</span><a class="reference internal" href="modules/seldon/pipeline/sklearn_transform.html#SklearnTransform"><span class="viewcode-link">[source]</span></a><a class="headerlink" href="#seldon.pipeline.sklearn_transform.SklearnTransform" title="Permalink to this definition">¶</a></dt>
<dd><p>Bases: <code class="xref py py-class docutils literal"><span class="pre">sklearn.base.BaseEstimator</span></code>, <code class="xref py py-class docutils literal"><span class="pre">sklearn.base.TransformerMixin</span></code></p>
<p>Allow sklearn transformers to be run on Pandas dataframes.</p>
<table class="docutils field-list" frame="void" rules="none">
<col class="field-name" />
<col class="field-body" />
<tbody valign="top">
<tr class="field-odd field"><th class="field-name">Parameters:</th><td class="field-body"><ul class="first last simple">
<li><strong>input_features</strong> (<em>list str</em>) &#8211; input columns to use</li>
<li><strong>output_features</strong> (<em>list str, optional</em>) &#8211; names of output columns</li>
<li><strong>transformer</strong> (<em>scikit learn Transformer</em>) &#8211; transformer to run on data</li>
<dl class="method">
<dt id="seldon.pipeline.sklearn_transform.SklearnTransform.fit">
<code class="descname">fit</code><span class="sig-paren">(</span><em>df</em><span class="sig-paren">)</span><a class="reference internal" href="modules/seldon/pipeline/sklearn_transform.html#SklearnTransform.fit"><span class="viewcode-link">[source]</span></a><a class="headerlink" href="#seldon.pipeline.sklearn_transform.SklearnTransform.fit" title="Permalink to this definition">¶</a></dt>
<dd><table class="docutils field-list" frame="void" rules="none">
<col class="field-name" />
<col class="field-body" />
<tbody valign="top">
<tr class="field-odd field"><th class="field-name">Parameters:</th><td class="field-body"><strong>df</strong> (<em>pandas dataframe</em>) &#8211; </td>
<tr class="field-even field"><th class="field-name">Returns:</th><td class="field-body"><strong>self</strong></td>
<tr class="field-odd field"><th class="field-name">Return type:</th><td class="field-body">object</td>
<dl class="method">
<dt id="seldon.pipeline.sklearn_transform.SklearnTransform.transform">
<code class="descname">transform</code><span class="sig-paren">(</span><em>df</em><span class="sig-paren">)</span><a class="reference internal" href="modules/seldon/pipeline/sklearn_transform.html#SklearnTransform.transform"><span class="viewcode-link">[source]</span></a><a class="headerlink" href="#seldon.pipeline.sklearn_transform.SklearnTransform.transform" title="Permalink to this definition">¶</a></dt>
<dd><p>transform the input columns and merge result into input dataframe using column names if provided</p>
<table class="docutils field-list" frame="void" rules="none">
<col class="field-name" />
<col class="field-body" />
<tbody valign="top">
<tr class="field-odd field"><th class="field-name">Parameters:</th><td class="field-body"><strong>df</strong> (<em>pandas dataframe</em>) &#8211; </td>
<tr class="field-even field"><th class="field-name">Returns:</th><td class="field-body"></td>
<tr class="field-odd field"><th class="field-name">Return type:</th><td class="field-body">Transformed pandas dataframe</td>
<div class="section" id="module-seldon.pipeline.tfidf_transform">
<span id="seldon-pipeline-tfidf-transform-module"></span><h2>seldon.pipeline.tfidf_transform module<a class="headerlink" href="#module-seldon.pipeline.tfidf_transform" title="Permalink to this headline">¶</a></h2>
<dl class="class">
<dt id="seldon.pipeline.tfidf_transform.TfidfTransform">
<em class="property">class </em><code class="descclassname">seldon.pipeline.tfidf_transform.</code><code class="descname">TfidfTransform</code><span class="sig-paren">(</span><em>input_feature=None, output_feature=None, min_df=0, max_df=1.0, select_features=False, topn_features=50000, stop_words=None, target_feature=None, vectorizer=None, TfidfTransformer=None, ch2=None, fnames=None, feature_names_support=[], ngram_range=[1, 1]</em><span class="sig-paren">)</span><a class="reference internal" href="modules/seldon/pipeline/tfidf_transform.html#TfidfTransform"><span class="viewcode-link">[source]</span></a><a class="headerlink" href="#seldon.pipeline.tfidf_transform.TfidfTransform" title="Permalink to this definition">¶</a></dt>
<dd><p>Bases: <code class="xref py py-class docutils literal"><span class="pre">sklearn.base.BaseEstimator</span></code>, <code class="xref py py-class docutils literal"><span class="pre">sklearn.base.TransformerMixin</span></code></p>
<p>Create TF-IDF (term frequency - inverse document frequency) features.</p>
<p>can use chi-squared test to limit features. Assumes string based input feature that can be split.
Uses scikit-learn based transformers internally</p>
<table class="docutils field-list" frame="void" rules="none">
<col class="field-name" />
<col class="field-body" />
<tbody valign="top">
<tr class="field-odd field"><th class="field-name">Parameters:</th><td class="field-body"><ul class="first last simple">
<li><strong>min_df</strong> (<em>int, optinal</em>) &#8211; min document frequency (for sklearn vectorizer)</li>
<li><strong>max_df</strong> (<em>float, optional</em>) &#8211; max document frequency (for sklearn vectorizer)</li>
<li><strong>select_features</strong> (<em>bool, optional</em>) &#8211; use chi-squared test to select features</li>
<li><strong>topn_features</strong> (<em>int, optional</em>) &#8211; keep top features from chi-squared test</li>
<li><strong>stop_words</strong> (<em>str, optional</em>) &#8211; stop words (for sklearn vectorizer)</li>
<li><strong>target_feature</strong> (<em>str, optional</em>) &#8211; target feature for chi-squared test</li>
<dl class="method">
<dt id="seldon.pipeline.tfidf_transform.TfidfTransform.fit">
<code class="descname">fit</code><span class="sig-paren">(</span><em>df</em><span class="sig-paren">)</span><a class="reference internal" href="modules/seldon/pipeline/tfidf_transform.html#TfidfTransform.fit"><span class="viewcode-link">[source]</span></a><a class="headerlink" href="#seldon.pipeline.tfidf_transform.TfidfTransform.fit" title="Permalink to this definition">¶</a></dt>
<dd><p>Fit tfidf transform</p>
<table class="docutils field-list" frame="void" rules="none">
<col class="field-name" />
<col class="field-body" />
<tbody valign="top">
<tr class="field-odd field"><th class="field-name">Parameters:</th><td class="field-body"><strong>df</strong> (<em>pandas dataframe</em>) &#8211; </td>
<tr class="field-even field"><th class="field-name">Returns:</th><td class="field-body"><strong>self</strong></td>
<tr class="field-odd field"><th class="field-name">Return type:</th><td class="field-body">object</td>
<dl class="method">
<dt id="seldon.pipeline.tfidf_transform.TfidfTransform.get_tokens">
<code class="descname">get_tokens</code><span class="sig-paren">(</span><em>v</em><span class="sig-paren">)</span><a class="reference internal" href="modules/seldon/pipeline/tfidf_transform.html#TfidfTransform.get_tokens"><span class="viewcode-link">[source]</span></a><a class="headerlink" href="#seldon.pipeline.tfidf_transform.TfidfTransform.get_tokens" title="Permalink to this definition">¶</a></dt>
<dd><p>basic method to get &#8220;document&#8221; string from feature</p>
<dl class="method">
<dt id="seldon.pipeline.tfidf_transform.TfidfTransform.transform">
<code class="descname">transform</code><span class="sig-paren">(</span><em>df</em><span class="sig-paren">)</span><a class="reference internal" href="modules/seldon/pipeline/tfidf_transform.html#TfidfTransform.transform"><span class="viewcode-link">[source]</span></a><a class="headerlink" href="#seldon.pipeline.tfidf_transform.TfidfTransform.transform" title="Permalink to this definition">¶</a></dt>
<dd><p>transform features with tfidf transform</p>
<table class="docutils field-list" frame="void" rules="none">
<col class="field-name" />
<col class="field-body" />
<tbody valign="top">
<tr class="field-odd field"><th class="field-name">Parameters:</th><td class="field-body"><strong>X</strong> (<em>pandas dataframe</em>) &#8211; </td>
<tr class="field-even field"><th class="field-name">Returns:</th><td class="field-body"></td>
<tr class="field-odd field"><th class="field-name">Return type:</th><td class="field-body">Transformed pandas dataframe</td>
<div class="section" id="module-seldon.pipeline">
<span id="module-contents"></span><h2>Module contents<a class="headerlink" href="#module-seldon.pipeline" title="Permalink to this headline">¶</a></h2>
           </div>
          </div>
          <footer>
    <div class="rst-footer-buttons" role="navigation" aria-label="footer navigation">
        <a href="seldon.text.html" class="btn btn-neutral float-right" title="seldon.text package" accesskey="n">Next <span class="fa fa-arrow-circle-right"></span></a>
        <a href="seldon.misc.html" class="btn btn-neutral" title="seldon.misc package" accesskey="p"><span class="fa fa-arrow-circle-left"></span> Previous</a>
  <div role="contentinfo">
        &copy; Copyright 2015, Seldon.
  Built with <a href="http://sphinx-doc.org/">Sphinx</a> using a <a href="https://github.com/snide/sphinx_rtd_theme">theme</a> provided by <a href="https://readthedocs.org">Read the Docs</a>. 
        </div>
      </div>
    </section>
    <script type="text/javascript">
        var DOCUMENTATION_OPTIONS = {
            URL_ROOT:'./',
            VERSION:'2.2.1',
            COLLAPSE_INDEX:false,
            FILE_SUFFIX:'.html',
            HAS_SOURCE:  true
    </script>
      <script type="text/javascript" src="static/jquery.js"></script>
      <script type="text/javascript" src="static/underscore.js"></script>
      <script type="text/javascript" src="static/doctools.js"></script>
    <script type="text/javascript" src="static/js/theme.js"></script>
  <script type="text/javascript">
      jQuery(function () {
          SphinxRtdTheme.StickyNav.enable();
  </script>
Provide feedback

Saved searches

Use saved searches to filter your results more quickly

FilesExpand file tree

seldon.pipeline.html

Latest commit

History

seldon.pipeline.html

File metadata and controls
