Package it.unimi.dsi.sux4j.io

Class BucketedHashStore<T>

java.lang.Object
it.unimi.dsi.sux4j.io.BucketedHashStore<T>
All Implemented Interfaces:
SafelyCloseable, Closeable, Serializable, AutoCloseable, Iterable<BucketedHashStore.Bucket>
public class BucketedHashStore<T> extends Object implements Serializable, SafelyCloseable, Iterable<BucketedHashStore.Bucket>
A temporary store of signatures virtually divided into buckets.

A bucketed hash store accumulates elements (objects of type T) by turning them into bit vectors (using a provided TransformationStrategy) and then hashing such vectors into a signature (a pair of longs, i.e., overall we get a hash of 128 bits). Elements can be added one by one or in batches. Elements must be distinct, or, more precisely, they must be transformed into distinct bit vectors.

Besides the hashes, we store some data associated with each element: if no data is specified, we store the rank of each element added (the first element added has rank 0, the second one has rank 1, and so on), unless you specified at construction time a nonzero hash width: in that case, the value stored by add(Object) will be given the lowest bits of the first hash of the signature associated with the object (the hash width is the number of bits stored). This feature makes it possible, for example, to implement a static dictionary using a GOV3Function.

The desired expected bucket size can be set by calling bucketSize(int). Once all elements have been added, one calls iterator(), which returns buckets one at a time (in their natural order); signatures within each bucket are returned by increasing value, and signatures within different buckets are in bucket order. Actually, the iterator provided by a bucket returns a triple of longs whose last element is the data associated with the element that generated the signature.

Note that the main difference between an instance of this class and one of a ChunkedHashStore is that the latter can only guarantee that the average chunk (here, bucket) size will be within a factor of two from the desired one, as the number of chunks can be specified only as a power of two.

It is possible (albeit very unlikely) that different elements generate the same hash. This event is detected during bucket iteration (not while accumulating hashes), and it will throw a BucketedHashStore.DuplicateException. At that point, the caller must handle the exception by resetting the store and trying again from scratch. Note that after a few (say, three) exceptions you can safely assume that there are duplicate elements. If you need to force a check on the whole store you can call check(). If all your elements come from an Iterable, checkAndRetry(Iterable, LongIterable) will try three times to build a checked bucketed hash store.

Every reset(long) changes the seed used by the store to generate signatures. So, if this seed has to be stored this must happen after the last call to reset(long). To help tracking this fact, a call to seed() will lock the store; any further call to reset(long) will throw an IllegalStateException. In case the store needs to be reused, you can call clear(), that will bring back the store to after-creation state.

When you have finished using a bucketed hash store, you should close() it. This class implements SafelyCloseable, and thus provides a safety-net finalizer.

Filtering

You can at any time set a predicate that will filter the signatures returned by the store.

Computing frequencies

If you specify so at construction time, a bucketed hash store will compute for you a a map from values to their frequency.

Implementation details

Internally, a bucketed hash store save signatures into different disk segments using the highest bits (performing, in fact, the first phase of a bucket sort). Once the user chooses a bucket size, the store exhibits the data on disk by grouping disk segments or splitting them into buckets. This process is transparent to the user.

The assignment to a bucket happens conceptually by using the first 64-bit hash (shifted by one to the right, to avoid sign issues) to define a number α in the interval [0..1). Then, the bucket assigned is ⌊αm⌋, where m = 1 + size() / bucketSize() is the number of buckets. Conceptually, we are mapping α, a uniform random number in the unit interval, into ⌊αm⌋, a uniform integer number in the range [0..m), by inversion. As show below, the whole computation can be carried out using a fixed-point representation, as it has been done for pseudorandom number generators since the early days, using only a multiplication and shift. The assignment is monotone nondecreasing, which makes it possible to emit the buckets one at a time scanning the keys in sorted order.

Signatures have to be loaded into memory only segment by segment, so to be sorted and tested for uniqueness. As long as DISK_SEGMENTS is larger than eight, the store will need less than 0.75 bits per element of main memory. DISK_SEGMENTS can be increased arbitrarily at compile time, but each store will open DISK_SEGMENTS files at the same time. (For the same reason, it is strongly suggested that you close your stores as soon as you do not need them).

Intended usage

bucketed hash stores should be built by classes that need to manipulate elements in buckets of approximate given size without needing access to the elements themselves, but just to their signatures, a typical example being GOV3Function, which uses the signatures to compute a 3-hyperedge. Once a bucketed hash store is built, it can be passed on to further substructures, reducing greatly the computation time (as the original collection need not to be scanned again).

To compute the bucket corresponding to a given element, use 

 final long[] signature = new long[2];
 Hashes.spooky4(transform.toBitVector(key), seed, signature);
 final int bucket = Math.multiplyHigh(signature[0] >>> 1, (1 + n / bucketSize) << 1);
where seed is the store seed, n is the number of keys, and bucketSize is the provided bucket size.
Since:
5.0.0
Author:
Sebastiano Vigna
See Also:

  • Field Details

    • serialVersionUID

      public static final long serialVersionUID
      See Also:

    • DEFAULT_BUCKET_SIZE

      public static final int DEFAULT_BUCKET_SIZE
      The default size of the bucket at creation. From Sux4J 5.2.0 it has been raised (it used to be one) to avoid that calls to checkAndRetry(Iterable) on a newly created store are too expensive.
      See Also:

    • BUFFER_SIZE

      public static final int BUFFER_SIZE
      The size of the output buffers.
      See Also:

    • LOG2_DISK_SEGMENTS

      public static final int LOG2_DISK_SEGMENTS
      The logarithm of the number of disk segments.
      See Also:

    • DISK_SEGMENTS

      public static final int DISK_SEGMENTS
      The number of disk segments.
      See Also:

    • DISK_SEGMENTS_SHIFT

      public static final int DISK_SEGMENTS_SHIFT
      The shift for disk segments.
      See Also:

    • size

      protected long size
      The number of elements ever added.

    • filteredSize

      protected long filteredSize
      The number of elements that pass the current filter, or -1 we it must be recomputed.

    • seed

      protected long seed
      The seed used to generate the hash signatures.

  • Constructor Details

    • BucketedHashStore

      public BucketedHashStore(TransformationStrategy<? super T> transform) throws IOException
      Creates a bucketed hash store with given transformation strategy.
      Parameters:
      transform - a transformation strategy for the elements.
      Throws:
      IOException

    • BucketedHashStore

      public BucketedHashStore(TransformationStrategy<? super T> transform, File tempDir) throws IOException
      Creates a bucketed hash store with given transformation strategy and temporary file directory.
      Parameters:
      transform - a transformation strategy for the elements.
      tempDir - a temporary directory for the store files, or null for the current directory.
      Throws:
      IOException

    • BucketedHashStore

      public BucketedHashStore(TransformationStrategy<? super T> transform, ProgressLogger pl) throws IOException
      Creates a bucketed hash store with given transformation strategy.
      Parameters:
      transform - a transformation strategy for the elements.
      pl - a progress logger, or null.
      Throws:
      IOException

    • BucketedHashStore

      public BucketedHashStore(TransformationStrategy<? super T> transform, File tempDir, ProgressLogger pl) throws IOException
      Creates a bucketed hash store with given transformation strategy and progress logger.
      Parameters:
      transform - a transformation strategy for the elements.
      tempDir - a temporary directory for the store files, or null for the current directory.
      pl - a progress logger, or null.
      Throws:
      IOException

    • BucketedHashStore

      public BucketedHashStore(TransformationStrategy<? super T> transform, File tempDir, int hashWidthOrCountValues, ProgressLogger pl) throws IOException
      Creates a bucketed hash store with given transformation strategy, hash width and progress logger.
      Parameters:
      transform - a transformation strategy for the elements.
      tempDir - a temporary directory for the store files, or null for the current directory.
      hashWidthOrCountValues - if positive, no associated data is saved in the store: BucketedHashStore.Bucket.data(long) will return this many lower bits of the first of the three hashes associated with the key; zero, values are stored; if negative, values are stored and a map from values to their frequency is computed.
      pl - a progress logger, or null.
      Throws:
      IOException

  • Method Details

    • bucketSize

      public int bucketSize()
      Returns the expected bucket size.
      Returns:
      the expected bucket size.

    • bucketSize

      public void bucketSize(int bucketSize)
      Sets the expected bucket size.
      Parameters:
      bucketSize - the expected bucket size.

    • seed

      public long seed()
      Return the current seed of this bucketed hash store. After calling this method, no reset(long) will be allowed (unless the store is cleared).
      Returns:
      the current seed of this bucketed hash store.

    • tempDir

      public File tempDir()
      Return the temporary directory of this bucketed hash store, or null.
      Returns:
      the temporary directory of this bucketed hash store, or null.

    • transform

      public TransformationStrategy<? super T> transform()
      Return the transformation strategy provided at construction time.
      Returns:
      the transformation strategy provided at construction time.

    • add

      public void add(T o, long value) throws IOException
      Adds an element to this store, associating it with a specified value.
      Parameters:
      o - the element to be added.
      value - the associated value.
      Throws:
      IOException

    • add

      public void add(T o) throws IOException
      Adds an element to this store, associating it with its ordinal position.
      Parameters:
      o - the element to be added.
      Throws:
      IOException

    • addAll

      public void addAll(Iterator<? extends T> elements, LongIterator values, boolean requiresValue2CountMap) throws IOException
      Adds the elements returned by an iterator to this store, associating them with specified values, possibly building the associated value frequency map.
      Parameters:
      elements - an iterator returning elements.
      values - an iterator on values parallel to elements.
      requiresValue2CountMap - whether to build the value frequency map (associating with each value its frequency).
      Throws:
      IOException

    • addAll

      public void addAll(Iterator<? extends T> elements, LongIterator values) throws IOException
      Adds the elements returned by an iterator to this store, associating them with specified values.
      Parameters:
      elements - an iterator returning elements.
      values - an iterator on values parallel to elements.
      Throws:
      IOException

    • addAll

      public void addAll(Iterator<? extends T> elements) throws IOException
      Adds the elements returned by an iterator to this store, associating them with their ordinal position.
      Parameters:
      elements - an iterator returning elements.
      Throws:
      IOException

    • size

      public long size() throws IOException
      Returns the size of this store. Note that if you set up a filter, the first call to this method will require a scan to the whole store.
      Returns:
      the number of (possibly filtered) pairs of this store.
      Throws:
      IOException

    • clear

      public void clear() throws IOException
      Clears this store. After a call to this method, the store can be reused.
      Throws:
      IOException

    • value2FrequencyMap

      public Long2LongOpenHashMap value2FrequencyMap()
      Return the current value frequency map.
      Returns:
      the current value frequency map.
      Throws:
      IllegalStateException - if this bucketed hash store does not contain a value frequency map.

    • finalize

      protected void finalize() throws Throwable
      Overrides:
      finalize in class Object
      Throws:
      Throwable

    • close

      public void close() throws IOException
      Closes this store, disposing all associated resources.
      Specified by:
      close in interface AutoCloseable
      Specified by:
      close in interface Closeable
      Throws:
      IOException

    • reset

      public void reset(long seed) throws IOException
      Resets this store using a new seed. All accumulated data are cleared, and a new seed is reinstated.
      Parameters:
      seed - the new seed.
      Throws:
      IllegalStateException - if this store was locked by a call to seed(), and never cleared thereafter.
      IOException

    • check

      public void check() throws BucketedHashStore.DuplicateException
      Checks that this store has no duplicate signatures, throwing an exception if this fails to happen.
      Throws:
      BucketedHashStore.DuplicateException - if this store contains duplicate signatures.

    • checkAndRetry

      public void checkAndRetry(Iterable<? extends T> iterable, LongIterable values) throws IOException
      Checks that this store has no duplicate signatures, and try to rebuild if this fails to happen.
      Parameters:
      iterable - the elements with which the store will be refilled if there are duplicate signatures.
      values - the values that will be associated with the elements returned by iterable.
      Throws:
      IllegalArgumentException - if after a few trials the store still contains duplicate signatures.
      IOException

    • checkAndRetry

      public void checkAndRetry(Iterable<? extends T> iterable) throws IOException
      Checks that this store has no duplicate signatures, and try to rebuild if this fails to happen.

      Warning: the actions are executed exactly in the specified order—first check, then retry. If you invoke this method on an empty store you'll get a checked empty store.

      Parameters:
      iterable - the elements with which the store will be refilled if there are duplicate signatures.
      Throws:
      IllegalArgumentException - if after a few trials the store still contains duplicate signatures.
      IOException

    • signatures

      public LongBigList signatures(int signatureWidth, ProgressLogger pl) throws IOException
      Generate a list of signatures using the lowest bits of the first hash in this store.

      For this method to work, this store must contain ranks.

      Parameters:
      signatureWidth - the width in bits of the signatures.
      pl - a progress logger.
      Throws:
      IOException

    • filter

      public void filter(Predicate<long[]> filter)
      Sets a filter for this store.
      Parameters:
      filter - a predicate that will be used to filter signatures.

    • iterator

      public Iterator<BucketedHashStore.Bucket> iterator()
      Returns an iterator over the buckets of this bucketed hash store.

      Note that at each iteration part of the state of this bucketed hash store is reused. Thus, after each call to next() the previously returned BucketedHashStore.Bucket will be no longer valid. Please use the provided copy constructor if you need to process in parallel several buckets.

      Specified by:
      iterator in interface Iterable<T>
      Returns:
      an iterator over the buckets of this bucketed hash store.