Class NumericRangeQuery<T>

A Query that matches numeric values within a specified range. To use this, you must first index the numeric values using NumericField (expert: NumericTokenStream ). If your terms are instead textual, you should use TermRangeQuery. NumericRangeFilter<T> is the filter equivalent of this query.

You create a new NumericRangeQuery with the static factory methods, eg:

Query q = NumericRangeQuery.newFloatRange("weight",
new Float(0.3f), new Float(0.10f),
true, true);

The performance of NumericRangeQuery is much better than the corresponding TermRangeQuery because the number of terms that must be searched is usually far fewer, thanks to trie indexing, described below.

You can optionally specify a precisionStep when creating this query. This is necessary if you've changed this configuration from its default (4) during indexing. Lower values consume more disk space but speed up searching. Suitable values are between 1 and 8. A good starting point to test is 4, which is the default value for all Numeric* classes. See below for details.

This query defaults to CONSTANT_SCORE_AUTO_REWRITE_DEFAULT for 32 bit (int/float) ranges with precisionStep <8 and 64 bit (long/double) ranges with precisionStep <6. Otherwise it uses CONSTANT_SCORE_FILTER_REWRITE as the number of terms is likely to be high. With precision steps of <4, this query can be run with one of the BooleanQuery rewrite methods without changing BooleanQuery's default max clause count.

NOTE: This API is experimental and might change in incompatible ways in the next release.

How it works

See the publication about panFMP, where this algorithm was described (referred to as TrieRangeQuery):

Schindler, U, Diepenbroek, M, 2008. Generic XML-based Framework for Metadata Portals. Computers & Geosciences 34 (12), 1947-1955. doi:10.1016/j.cageo.2008.02.023

A quote from this paper: Because Apache Lucene is a full-text search engine and not a conventional database, it cannot handle numerical ranges (e.g., field value is inside user defined bounds, even dates are numerical values). We have developed an extension to Apache Lucene that stores the numerical values in a special string-encoded format with variable precision (all numerical values like doubles, longs, floats, and ints are converted to lexicographic sortable string representations and stored with different precisions (for a more detailed description of how the values are stored, see Lucene.Net.Util.NumericUtils). A range is then divided recursively into multiple intervals for searching: The center of the range is searched only with the lowest possible precision in the trie, while the boundaries are matched more exactly. This reduces the number of terms dramatically.

For the variant that stores long values in 8 different precisions (each reduced by 8 bits) that uses a lowest precision of 1 byte, the index contains only a maximum of 256 distinct values in the lowest precision. Overall, a range could consist of a theoretical maximum of 72552 + 255 = 3825 distinct terms (when there is a term for every distinct value of an 8-byte-number in the index and the range covers almost all of them; a maximum of 255 distinct values is used because it would always be possible to reduce the full 256 values to one term with degraded precision). In practice, we have seen up to 300 terms in most cases (index with 500,000 metadata records and a uniform value distribution).

Precision Step

You can choose any precisionStep when encoding values. Lower step values mean more precisions and so more terms in index (and index gets larger). On the other hand, the maximum number of terms to match reduces, which optimized query speed. The formula to calculate the maximum term count is:

n = [ (bitsPerValue/precisionStep - 1) * (2^precisionStep - 1 ) * 2 ] + (2^precisionStep - 1 )

(this formula is only correct, when bitsPerValue/precisionStep is an integer; in other cases, the value must be rounded up and the last summand must contain the modulo of the division as precision step). For longs stored using a precision step of 4, n = 15152 + 15 = 465, and for a precision step of 2, n = 3132 + 3 = 189. But the faster search speed is reduced by more seeking in the term enum of the index. Because of this, the ideal precisionStep value can only be found out by testing. Important: You can index with a lower precision step value and test search speed using a multiple of the original step value.

Good values for precisionStep are depending on usage and data type:

Comparisons of the different types of RangeQueries on an index with about 500,000 docs showed that TermRangeQuery in boolean rewrite mode (with raised BooleanQuery clause count) took about 30-40 secs to complete, TermRangeQuery in constant score filter rewrite mode took 5 secs and executing this class took <100ms to complete (on an Opteron64 machine, Java 1.5, 8 bit precision step). This query type was developed for a geographic portal, where the performance for e.g. bounding boxes or exact date/time stamps is important.