picky

Picky::Index#ranged_category

Make this category range searchable with a fixed range. If you need other
ranges, define another category with a different range value.

Example:
You have data values inside 1..100, and you want to have Picky return
not only the results for 47 if you search for 47, but also results for
45, 46, or 47.2, 48.9, in a range of 2 around 47, so (45..49).

Then you use:
 ranged_category :values_inside_1_100, 2

Optionally, you give it a precision value to reduce the error margin
around 47 (Picky is a bit liberal).
  Index.new :range do
    ranged_category :values_inside_1_100, 2, precision: 5
  end

This will force Picky to maximally be wrong 5% of the given range value
(5% of 2 = 0.1) instead of the default 20% (20% of 2 = 0.4).

We suggest not to use much more than 5 as a higher precision is more
performance intensive for less and less precision gain.

== Protip 1

Create two ranged categories to make an area search:
  Index.new :area do
    ranged_category :x, 1
    ranged_category :y, 1
  end

Search for it using for example:
  x:133, y:120

This will search this square area (* = 133, 120: The "search" point entered):

   132       134
    |         |
  --|---------|-- 121
    |         |
    |    *    |
    |         |
  --|---------|-- 119
    |         |

Note: The area does not need to be square, but can be rectangular.

== Protip 2

Create three ranged categories to make a volume search.

Or go crazy and use 4 ranged categories for a space/time search! ;)

=== Parameters
* category_name: The category_name as used in #category.
* range: The range (in the units of your data values) around the query point where we search for results.

 -----|<- range  ->*------------|-----

=== Options
* precision: Default is 1 (20% error margin, very fast), up to 5 (5% error margin, slower) makes sense.
* anchor: Where to anchor the grid.
* ... all options of #category.

Source | Google | Stack overflow