The Bloom Filter Implementation in Golang with Redis.

• Bloom Filter Calculator
• Bloom Filter Calculator2

n: the number of items in filter. p: the probability of false positives. (0.01 means 1%) k: the number of hash functions. m: the number of bits in the filter.(the needed memory).

can use the link blow the estimate the n,p,k,m.

1. Can use the n,p as the input, then you can get hte k,m.

func EstimateParameters(n uint, p float64) (m int64, k int64)

2. Can use the n,p,k as the input, then you can get the m. For the solution, you can directly call

func New(m int64, k int64, bitSet BitSetProvider) *BloomFilter

The difference is you can set the k yourself or use the value calculated by the system.

var client *redis.Client
client = redis.NewClient(&redis.Options{
    Addr:     "127.0.0.1:6379",
    Password: "",
    DB:       0,
    PoolSize: 256,
})
//use the estimate m and k.
m, k := bloom.EstimateParameters(100000, 0.001)
//new a Bloom Filter
bitSet := bloom.NewRedisBitSet("test_key", m, client)
b := bloom.New(m, k, bitSet)

//check exist
data := []byte("some key")
exists, err := b.Exists(data)
err = b.Add(data)
exists, err := b.Exists(data)