Minor improvements

- State that it's the dense dimension that must be identical in closeness.
- Use "double".

Author: bratseth

en/reference/ranking/rank-features.html

@@ -18,17 +18,20 @@
 </p>
 <ul>
   <li>
-    Types: All rank feature values are floats. Integers are converted to exact whole value floats.
-    String values are converted to exact whole value floats using a hash function.
+    Types: All rank feature values are doubles or tensors. Integers are converted to exact whole value doubles.
+    String values are converted to exact whole value doubles using a hash function.
     String literals in ranking expressions are converted using the same hash function,
     to enable equality tests on string values.
-  </li><li>
+  </li>
+  <li>
     Features which are <em>normalized</em> are between 0 and 1, where 0 is always the minimum and 1 the maximum.
     Normalized features should normally be preferred because they are more easily combined by
     <a href="ranking-expressions.html">ranking expressions</a> into a complete normalized score.
-  </li><li>
+  </li>
+  <li>
     A query may override <em>any</em> rank feature value by submitting that value as a feature with the query.
-  </li><li>
+  </li>
+  <li>
     Some features have parameters. It is always allowed to quote parameters with <em>"</em>.
     Nested quotes are not allowed and must be escaped using <em>\</em>.
     Parameters that can be parsed as feature names may be left unquoted.
@@ -42,11 +45,7 @@ <h2 id="feature-list">Feature list</h2>
 <!--table-to-list-->
 <table class="table" id="rank-feature-table">
 
-
-
-
-
-<tr><td colspan="3"></td></tr>
+  <tr><td colspan="3"></td></tr>
 <tr><td colspan="3"><h3 id="query-features">Query features</h3></td></tr>
 <tr class="trx"><th class="thx">Feature name</th><th class="thx">Default</th><th class="thx">Description</th></tr>
 
@@ -381,7 +380,7 @@ <h2 id="feature-list">Feature list</h2>
   <td>0</td>
   <td><p id="fieldMatch(name).absoluteOccurrence">
     Returns a normalized measure of the number of occurrence of the terms of the query:
-    </p><p class="equation-container" class="equation-container"><!-- depends on mathjax -->
+    </p><p class="equation-container"><!-- depends on mathjax -->
     $$\frac{\sum_{\text{all query terms}}(min(\text{number of occurrences of the term},maxOccurrences))}{(\text{query term count} &times; 100)}$$
     </p><p>
     This is 1 if there are many occurrences of the query terms, and 0 if there are none.
@@ -838,10 +837,8 @@ <h3 id="attribute-match-features-normalized">Attribute match features - normaliz
   <td>0</td>
   <td><p id="closeness(dimension,name)">
     Used with the <a href="../querying/yql.html#nearestneighbor">nearestNeighbor</a> query operator.
-    A number which is close to 1 when a point vector in the document is close
-    to a matching point vector in the query.
-    The document vectors and the query vector must be the same tensor
-    type, with one indexed dimension of size N, representing a point in an N-dimensional space.
+    A number which is close to 1 when a vector in the document tensor is close to the vector given in the query.
+    The indexed dimension representing a vector in document and query tensors must be identical.
   </p>
     <ul>
       <li>
@@ -899,12 +896,12 @@ <h3 id="attribute-match-features-normalized">Attribute match features - normaliz
   <td><p id="freshness(name).logscale">
     A logarithmic-shaped freshness; also goes from 1 to 0, but looks like <a href="#freshness">freshness plot</a>.
     The function is based on <code>-log(age(name) + scale)</code> and is calculated as:
-    </p><p class="equation-container" class="equation-container"><!-- depends on mathjax -->
+    </p><p class="equation-container"><!-- depends on mathjax -->
     $$\frac{log(maxAge + scale) - log(age(name) + scale)}{log(maxAge + scale) - log(scale)}$$
     </p><p>
     where scale is defined using
     <a href="rank-feature-configuration.html#freshness"> halfResponse and maxAge</a>:
-    </p><p class="equation-container" class="equation-container"><!-- depends on mathjax -->
+    </p><p class="equation-container"><!-- depends on mathjax -->
     $$\frac{-halfResponse^2}{2 &times; halfResponse - maxAge}$$
     </p><p>
     When <code>age(name) == halfResponse</code> the function output is 0.5.</p>