Merge pull request #31 from apache/sidebar-v2

Sidebar v2

Author: jmalkin

docs/source/cpc.rst docs/source/distinct_counting/cpc.rstdocs/source/cpc.rst renamed to docs/source/distinct_counting/cpc.rst

@@ -0,0 +1,24 @@
+Distinct Counting 
+=================
+
+.. currentmodule:: datasketches
+
+Distinct counting is one of the earliest tasks to which sketches were applied. The concept is simple:
+Provide an estimate of the number of unique elements in a set of data. One of the earliest solutions came
+from Flajolet and Martin in 1985 with their seminal work
+`Probabilistic counting Algorithms for Data Base Applications <http://db.cs.berkeley.edu/cs286/papers/flajoletmartin-jcss1985.pdf>`_.
+
+The DataSketches library offers several types of distinct counting sketches, each with different properties.
+
+  * :class:`hll_sketch`: Hyper Log Log, a well-known sketch for distinct counting but no longer state-of-the-art.
+  * :class:`cpc_sketch`: Provides a better accuracy-space trade-off than HLL, but with a somewhat larger footprint while in-memory.
+  * :class:`theta_sketch`: Theta sketch, a type of k-minimum value sketch, which provide good performance with intersection and set difference operations.
+  * :class:`tuple_sketch`: Tuple sketch, which is similar to a theta sketch but supports additional data stored with each key.
+
+.. toctree::
+  :maxdepth: 1
+   
+  hyper_log_log
+  cpc
+  theta 
+  tuple

docs/source/hyper_log_log.rst …urce/distinct_counting/hyper_log_log.rstdocs/source/hyper_log_log.rst renamed to docs/source/distinct_counting/hyper_log_log.rst docs/source/hyper_log_log.rst renamed to docs/source/distinct_counting/hyper_log_log.rst

@@ -0,0 +1,15 @@
+Frequency Sketches
+==================
+
+Frequency estimation involves determining how often an item has been seen in a stream. The library currently
+offers two types of sketches for frequency estimation, one of which has two closely-related variants.
+
+  * :class:`frequent_items_sketch`: Identifies the *Top K* or *heavy hitters* in a stream, those items whose weight is above a certain percentage of the entire stream. Does not necessarily provide an estimate for most items outside the heavy hitters.
+  * :class:`frequent_strings_sketch`: Like the items version but containing snly strings (an implementation from before the library handled generic objects).
+  * :class:`count_min_sketch`: Provides an estimate for any item, regardless of relative weight, but does not maintain a list of the heaviest items.
+
+.. toctree::
+  :maxdepth: 1
+   
+  frequent_items
+  count_min_sketch

docs/source/distinct_counting/index.rst

@@ -0,0 +1,24 @@
+Helper Classes
+==============
+
+.. currentmodule:: dataksetches
+
+These classes are required for certain sketches or specific
+functionality within sketches.
+Some of them are abstract base classes, but in those cases there is at
+least one reference example of a concrete class.
+
+  * :class:`serde` is used when serializing and deserializing sketches.
+  * :class:`jaccard` is used to compute the Jaccard similarity between pairs of theta or tuple sketches.
+  * :class:`tuple_policy` is required to use a :class:`tuple_sketch` by specifying how summaries are combined.
+  * :func:`ks_test` performs a Kolmogorov-Smirnov test on absolute-error quantiles family sketches.
+  * :class:`kernel_function` is required when using a :class:`kernel_sketch` for Kernel Density Estimation.
+
+.. toctree::
+  :maxdepth: 1
+
+  serde
+  jaccard
+  tuple_policy
+  ks_test
+  kernel