bloom-filter

  1. bloom-filter
  2. Documentation
    1. make-bloom-filter
    2. bloom-filter-set!
    3. bloom-filter-exists?
    4. bloom-filter?
    5. check-bloom-filter
    6. error-bloom-filter
    7. bloom-filter-algorithms
    8. bloom-filter-n
    9. bloom-filter-m
    10. bloom-filter-k
    11. bloom-filter-p-false-positive
    12. actual-k
    13. optimum-size
    14. desired-m
    15. optimum-k
    16. optimum-m
    17. p-false-positive
    18. p-random-one-bit
  3. Usage
  4. References
  5. Requirements
  6. Author
  7. Repository
  8. Version history
  9. License

Documentation

Provides a simple Bloom Filter

Bloom Filter Object

make-bloom-filter

[procedure] (make-bloom-filter M MDPS [K]) --> bloom-filter

Returns a bloom-filter object with M bits of discrimination and a set of hash functions built from the supplied MDPS, a (list-of message-digest-primitive) objects.

The number of hashes, K, is not necessarily the same as the number of message-digests. A hash (here) is defined as an unsigned 32/64 bit integer. Most message-digests return more 32/64 bits of hash. The actual length of the hash is divided into 32/64 bit blocks to get the individual hashes.

The argument K will restrict the actual number of hashes to the "first" K, no matter how many more the supplied message-digests create. First in the order of MDPS.

[procedure] (make-bloom-filter P N MDPS) --> bloom-filter

Returns a bloom-filter object with M and K values chosen for the given population capacity N and probablity of false-positives P.

Selecting the optimal set of message-digests is beyond the scope of make-bloom-filter.

bloom-filter-set!

[procedure] (bloom-filter-set! BLOOM-FILTER OBJECT)

Add the specified OBJECT to the indicated BLOOM-FILTER.

bloom-filter-exists?

[procedure] (bloom-filter-exists? BLOOM-FILTER OBJECT) -> boolean

Is the specified OBJECT in the indicated BLOOM-FILTER.

bloom-filter?

check-bloom-filter

error-bloom-filter

[procedure] (bloom-filter? OBJ) --> boolean
[procedure] (check-bloom-filter LOC OBJ [NAM]) --> bloom-filter
[procedure] (error-bloom-filter LOC OBJ [NAM])

bloom-filter-algorithms

[procedure] (bloom-filter-algorithms BLOOM-FILTER) --> (list-of message-digest-primitive)

The mdps used for the filter.

bloom-filter-n

[procedure] (bloom-filter-n BLOOM-FILTER) -> fixnum

The current population - the number of objects added to the filter.

Not the population capacity.

bloom-filter-m

[procedure] (bloom-filter-m BLOOM-FILTER) --> fixnum

The number of bits of discrimination.

bloom-filter-k

[procedure] (bloom-filter-k BLOOM-FILTER) --> fixnum

The number of hashes. (See above.)

bloom-filter-p-false-positive

[procedure] (bloom-filter-p-false-positive BLOOM-FILTER [N]) -> float

The probability of a false-positive for the population capacity N, default is the current population, bloom-filter-n.

actual-k

[procedure] (actual-k MDPS) --> fixnum

Calculates the actual number of hashes for the MDPS.

optimum-size

[procedure] (optimum-size P N) --> fixnum fixnum

Returns 2 values, an optimal M, bits of discrimination, and K, number of hashes, for the given population size N and probability of false-positives P.

desired-m

[procedure] (desired-m P N [K]) --> fixnum fixnum float

Calculates a near-optimal number of bits of discrimination to meet the desired probability of false positives P, with the given population size N and number of hashes K. When the K parameter is missing optimum-k is used to calculate a value.

A multi-valued return of the calculated M, K, and P values. The calculated probability may be lower than the desired. The calculated M value will always be a fixnum.

optimum-k

[procedure] (optimum-k N M) --> fixnum

Optimal count of hashes for the given population size N and M bits of discrimination.

optimum-m

[procedure] (optimum-m K N) --> fixnum

Optimal count of bits of discrimination for the given population size N and K number of hashes.

p-false-positive

[procedure] (p-false-positive K N M) --> float

What is the probability of false positives for the population size N assuming K hashes and M bits of discrimination.

p-random-one-bit

[procedure] (p-random-one-bit K N M) --> float

Calculates the probablility of a random set bit for the given number of hash functions K, population size N, and bits of discrimination M.

Usage

(import bloom-filter)

References

Nice exposition of Bloom Filter False Positive Probability.

Requirements

iset message-digest-primitive message-digest-type message-digest-utils check-errors

test test-utils sha1 md5 sha2 tiger-hash ripemd

Author

Kon Lovett

Repository

This egg is hosted on the CHICKEN Subversion repository:

https://anonymous@code.call-cc.org/svn/chicken-eggs/release/5/bloom-filter

If you want to check out the source code repository of this egg and you are not familiar with Subversion, see this page.

Version history

2.3.3
Move operation into C.
2.3.2
Smaller getter.
2.3.1
Fix datum getter (C).
2.3.0
Fix actual-k; use fixed wordsize for all machines.
2.2.9
.
2.2.2
Fix bloom-filter-p-false-positive, bloom-filter-n & bloom-filter-exists? signature.
2.2.0
Remove record-variants requirement.
2.0.0
CHICKEN 5 release.

License

Copyright (C) 2010-2023 Kon Lovett. All rights reserved.

Permission is hereby granted, free of charge, to any person obtaining a copy of this software and associated documentation files (the Software), to deal in the Software without restriction, including without limitation the rights to use, copy, modify, merge, publish, distribute, sublicense, and/or sell copies of the Software, and to permit persons to whom the Software is furnished to do so, subject to the following conditions:

The above copyright notice and this permission notice shall be included in all copies or substantial portions of the Software.

THE SOFTWARE IS PROVIDED ASIS, WITHOUT WARRANTY OF ANY KIND, EXPRESS OR IMPLIED, INCLUDING BUT NOT LIMITED TO THE WARRANTIES OF MERCHANTABILITY, FITNESS FOR A PARTICULAR PURPOSE AND NONINFRINGEMENT. IN NO EVENT SHALL THE AUTHORS OR COPYRIGHT HOLDERS BE LIABLE FOR ANY CLAIM, DAMAGES OR OTHER LIABILITY, WHETHER IN AN ACTION OF CONTRACT, TORT OR OTHERWISE, ARISING FROM, OUT OF OR IN CONNECTION WITH THE SOFTWARE OR THE USE OR OTHER DEALINGS IN THE SOFTWARE.