RDD
Related Docs:
object RDD
| package rdd
Construct an RDD with just a one-to-one dependency on one parent
:: DeveloperApi :: Implemented by subclasses to compute a given partition.
:: DeveloperApi :: Implemented by subclasses to compute a given partition.
Implemented by subclasses to return the set of partitions in this RDD.
Implemented by subclasses to return the set of partitions in this RDD. This method will only be called once, so it is safe to implement a time-consuming computation in it.
The partitions in this array must satisfy the following property:
rdd.partitions.zipWithIndex.forall { case (partition, index) => partition.index == index }
Return the union of this RDD and another one.
Return the union of this RDD and another one. Any identical elements will appear multiple
times (use .distinct() to eliminate them).
Aggregate the elements of each partition, and then the results for all the partitions, using given combine functions and a neutral "zero value".
Aggregate the elements of each partition, and then the results for all the partitions, using given combine functions and a neutral "zero value". This function can return a different result type, U, than the type of this RDD, T. Thus, we need one operation for merging a T into an U and one operation for merging two U's, as in scala.TraversableOnce. Both of these functions are allowed to modify and return their first argument instead of creating a new U to avoid memory allocation.
the initial value for the accumulated result of each partition for the
seqOp operator, and also the initial value for the combine results from
different partitions for the combOp operator - this will typically be the
neutral element (e.g. Nil for list concatenation or 0 for summation)
an operator used to accumulate results within a partition
an associative operator used to combine results from different partitions
Persist this RDD with the default storage level (MEMORY_ONLY).
Return the Cartesian product of this RDD and another one, that is, the RDD of all pairs of
elements (a, b) where a is in this and b is in other.
Mark this RDD for checkpointing.
Mark this RDD for checkpointing. It will be saved to a file inside the checkpoint
directory set with SparkContext#setCheckpointDir and all references to its parent
RDDs will be removed. This function must be called before any job has been
executed on this RDD. It is strongly recommended that this RDD is persisted in
memory, otherwise saving it on a file will require recomputation.
Clears the dependencies of this RDD.
Clears the dependencies of this RDD. This method must ensure that all references to the original parent RDDs are removed to enable the parent RDDs to be garbage collected. Subclasses of RDD may override this method for implementing their own cleaning logic. See org.apache.spark.rdd.UnionRDD for an example.
Return a new RDD that is reduced into numPartitions partitions.
Return a new RDD that is reduced into numPartitions partitions.
This results in a narrow dependency, e.g. if you go from 1000 partitions to 100 partitions, there will not be a shuffle, instead each of the 100 new partitions will claim 10 of the current partitions. If a larger number of partitions is requested, it will stay at the current number of partitions.
However, if you're doing a drastic coalesce, e.g. to numPartitions = 1, this may result in your computation taking place on fewer nodes than you like (e.g. one node in the case of numPartitions = 1). To avoid this, you can pass shuffle = true. This will add a shuffle step, but means the current upstream partitions will be executed in parallel (per whatever the current partitioning is).
With shuffle = true, you can actually coalesce to a larger number of partitions. This is useful if you have a small number of partitions, say 100, potentially with a few partitions being abnormally large. Calling coalesce(1000, shuffle = true) will result in 1000 partitions with the data distributed using a hash partitioner. The optional partition coalescer passed in must be serializable.
Return an RDD that contains all matching values by applying f.
Return an array that contains all of the elements in this RDD.
Return an array that contains all of the elements in this RDD.
This method should only be used if the resulting array is expected to be small, as all the data is loaded into the driver's memory.
The org.apache.spark.SparkContext that this RDD was created on.
Return the number of elements in the RDD.
Approximate version of count() that returns a potentially incomplete result within a timeout, even if not all tasks have finished.
Approximate version of count() that returns a potentially incomplete result within a timeout, even if not all tasks have finished.
The confidence is the probability that the error bounds of the result will contain the true value. That is, if countApprox were called repeatedly with confidence 0.9, we would expect 90% of the results to contain the true count. The confidence must be in the range [0,1] or an exception will be thrown.
maximum time to wait for the job, in milliseconds
the desired statistical confidence in the result
a potentially incomplete result, with error bounds
Return approximate number of distinct elements in the RDD.
Return approximate number of distinct elements in the RDD.
The algorithm used is based on streamlib's implementation of "HyperLogLog in Practice: Algorithmic Engineering of a State of The Art Cardinality Estimation Algorithm", available here.
Relative accuracy. Smaller values create counters that require more space. It must be greater than 0.000017.
Return approximate number of distinct elements in the RDD.
Return approximate number of distinct elements in the RDD.
The algorithm used is based on streamlib's implementation of "HyperLogLog in Practice: Algorithmic Engineering of a State of The Art Cardinality Estimation Algorithm", available here.
The relative accuracy is approximately 1.054 / sqrt(2^p). Setting a nonzero (sp is greater
than p) would trigger sparse representation of registers, which may reduce the memory
consumption and increase accuracy when the cardinality is small.
The precision value for the normal set.
p must be a value between 4 and sp if sp is not zero (32 max).
The precision value for the sparse set, between 0 and 32.
If sp equals 0, the sparse representation is skipped.
Return the count of each unique value in this RDD as a local map of (value, count) pairs.
Return the count of each unique value in this RDD as a local map of (value, count) pairs.
This method should only be used if the resulting map is expected to be small, as the whole thing is loaded into the driver's memory. To handle very large results, consider using
rdd.map(x => (x, 1L)).reduceByKey(_ + _)
, which returns an RDD[T, Long] instead of a map.
Approximate version of countByValue().
Approximate version of countByValue().
maximum time to wait for the job, in milliseconds
the desired statistical confidence in the result
a potentially incomplete result, with error bounds
Get the list of dependencies of this RDD, taking into account whether the RDD is checkpointed or not.
Return a new RDD containing the distinct elements in this RDD.
Return a new RDD containing the distinct elements in this RDD.
Return a new RDD containing only the elements that satisfy a predicate.
Return the first element in this RDD.
Returns the first parent RDD
Returns the first parent RDD
Return a new RDD by first applying a function to all elements of this RDD, and then flattening the results.
Aggregate the elements of each partition, and then the results for all the partitions, using a given associative function and a neutral "zero value".
Aggregate the elements of each partition, and then the results for all the partitions, using a given associative function and a neutral "zero value". The function op(t1, t2) is allowed to modify t1 and return it as its result value to avoid object allocation; however, it should not modify t2.
This behaves somewhat differently from fold operations implemented for non-distributed collections in functional languages like Scala. This fold operation may be applied to partitions individually, and then fold those results into the final result, rather than apply the fold to each element sequentially in some defined ordering. For functions that are not commutative, the result may differ from that of a fold applied to a non-distributed collection.
the initial value for the accumulated result of each partition for the op
operator, and also the initial value for the combine results from different
partitions for the op operator - this will typically be the neutral
element (e.g. Nil for list concatenation or 0 for summation)
an operator used to both accumulate results within a partition and combine results from different partitions
Applies a function f to all elements of this RDD.
Applies a function f to each partition of this RDD.
Gets the name of the directory to which this RDD was checkpointed.
Gets the name of the directory to which this RDD was checkpointed. This is not defined if the RDD is checkpointed locally.
Implemented by subclasses to return how this RDD depends on parent RDDs.
Implemented by subclasses to return how this RDD depends on parent RDDs. This method will only be called once, so it is safe to implement a time-consuming computation in it.
Returns the number of partitions of this RDD.
Returns the number of partitions of this RDD.
Optionally overridden by subclasses to specify placement preferences.
Optionally overridden by subclasses to specify placement preferences.
Get the RDD's current storage level, or StorageLevel.NONE if none is set.
Return an RDD created by coalescing all elements within each partition into an array.
Return an RDD of grouped items.
Return an RDD of grouped items. Each group consists of a key and a sequence of elements mapping to that key. The ordering of elements within each group is not guaranteed, and may even differ each time the resulting RDD is evaluated.
This operation may be very expensive. If you are grouping in order to perform an
aggregation (such as a sum or average) over each key, using PairRDDFunctions.aggregateByKey
or PairRDDFunctions.reduceByKey will provide much better performance.
Return an RDD of grouped elements.
Return an RDD of grouped elements. Each group consists of a key and a sequence of elements mapping to that key. The ordering of elements within each group is not guaranteed, and may even differ each time the resulting RDD is evaluated.
This operation may be very expensive. If you are grouping in order to perform an
aggregation (such as a sum or average) over each key, using PairRDDFunctions.aggregateByKey
or PairRDDFunctions.reduceByKey will provide much better performance.
Return an RDD of grouped items.
Return an RDD of grouped items. Each group consists of a key and a sequence of elements mapping to that key. The ordering of elements within each group is not guaranteed, and may even differ each time the resulting RDD is evaluated.
This operation may be very expensive. If you are grouping in order to perform an
aggregation (such as a sum or average) over each key, using PairRDDFunctions.aggregateByKey
or PairRDDFunctions.reduceByKey will provide much better performance.
A unique ID for this RDD (within its SparkContext).
Return the intersection of this RDD and another one.
Return the intersection of this RDD and another one. The output will not contain any duplicate elements, even if the input RDDs did. Performs a hash partition across the cluster
How many partitions to use in the resulting RDD
This method performs a shuffle internally.
Return the intersection of this RDD and another one.
Return the intersection of this RDD and another one. The output will not contain any duplicate elements, even if the input RDDs did.
Partitioner to use for the resulting RDD
This method performs a shuffle internally.
Return the intersection of this RDD and another one.
Return the intersection of this RDD and another one. The output will not contain any duplicate elements, even if the input RDDs did.
This method performs a shuffle internally.
Return whether this RDD is checkpointed and materialized, either reliably or locally.
true if and only if the RDD contains no elements at all. Note that an RDD may be empty even when it has at least 1 partition.
Due to complications in the internal implementation, this method will raise an
exception if called on an RDD of Nothing or Null. This may be come up in practice
because, for example, the type of parallelize(Seq()) is RDD[Nothing].
(parallelize(Seq()) should be avoided anyway in favor of parallelize(Seq[T]()).)
Internal method to this RDD; will read from cache if applicable, or otherwise compute it.
Internal method to this RDD; will read from cache if applicable, or otherwise compute it. This should not be called by users directly, but is available for implementors of custom subclasses of RDD.
Creates tuples of the elements in this RDD by applying f.
Mark this RDD for local checkpointing using Spark's existing caching layer.
Mark this RDD for local checkpointing using Spark's existing caching layer.
This method is for users who wish to truncate RDD lineages while skipping the expensive step of replicating the materialized data in a reliable distributed file system. This is useful for RDDs with long lineages that need to be truncated periodically (e.g. GraphX).
Local checkpointing sacrifices fault-tolerance for performance. In particular, checkpointed data is written to ephemeral local storage in the executors instead of to a reliable, fault-tolerant storage. The effect is that if an executor fails during the computation, the checkpointed data may no longer be accessible, causing an irrecoverable job failure.
This is NOT safe to use with dynamic allocation, which removes executors along
with their cached blocks. If you must use both features, you are advised to set
spark.dynamicAllocation.cachedExecutorIdleTimeout to a high value.
The checkpoint directory set through SparkContext#setCheckpointDir is not used.
Return a new RDD by applying a function to all elements of this RDD.
Return a new RDD by applying a function to each partition of this RDD.
Return a new RDD by applying a function to each partition of this RDD.
preservesPartitioning indicates whether the input function preserves the partitioner, which
should be false unless this is a pair RDD and the input function doesn't modify the keys.
Return a new RDD by applying a function to each partition of this RDD, while tracking the index of the original partition.
Return a new RDD by applying a function to each partition of this RDD, while tracking the index of the original partition.
preservesPartitioning indicates whether the input function preserves the partitioner, which
should be false unless this is a pair RDD and the input function doesn't modify the keys.
Returns the max of this RDD as defined by the implicit Ordering[T].
Returns the max of this RDD as defined by the implicit Ordering[T].
the maximum element of the RDD
Returns the min of this RDD as defined by the implicit Ordering[T].
Returns the min of this RDD as defined by the implicit Ordering[T].
the minimum element of the RDD
A friendly name for this RDD
Returns the jth parent RDD: e.g.
Returns the jth parent RDD: e.g. rdd.parent[T](0) is equivalent to rdd.firstParent[T]
Optionally overridden by subclasses to specify how they are partitioned.
Get the array of partitions of this RDD, taking into account whether the RDD is checkpointed or not.
Persist this RDD with the default storage level (MEMORY_ONLY).
Set this RDD's storage level to persist its values across operations after the first time it is computed.
Set this RDD's storage level to persist its values across operations after the first time it is computed. This can only be used to assign a new storage level if the RDD does not have a storage level set yet. Local checkpointing is an exception.
Return an RDD created by piping elements to a forked external process.
Return an RDD created by piping elements to a forked external process. The resulting RDD is computed by executing the given process once per partition. All elements of each input partition are written to a process's stdin as lines of input separated by a newline. The resulting partition consists of the process's stdout output, with each line of stdout resulting in one element of the output partition. A process is invoked even for empty partitions.
The print behavior can be customized by providing two functions.
command to run in forked process.
environment variables to set.
Before piping elements, this function is called as an opportunity to pipe context data. Print line function (like out.println) will be passed as printPipeContext's parameter.
Use this function to customize how to pipe elements. This function will be called with each RDD element as the 1st parameter, and the print line function (like out.println()) as the 2nd parameter. An example of pipe the RDD data of groupBy() in a streaming way, instead of constructing a huge String to concat all the elements:
def printRDDElement(record:(String, Seq[String]), f:String=>Unit) = for (e <- record._2) {f(e)}
Use separate working directories for each task.
Buffer size for the stdin writer for the piped process.
Char encoding used for interacting (via stdin, stdout and stderr) with the piped process
the result RDD
Return an RDD created by piping elements to a forked external process.
Return an RDD created by piping elements to a forked external process.
Get the preferred locations of a partition, taking into account whether the RDD is checkpointed.
Randomly splits this RDD with the provided weights.
Randomly splits this RDD with the provided weights.
weights for splits, will be normalized if they don't sum to 1
random seed
split RDDs in an array
Reduces the elements of this RDD using the specified commutative and associative binary operator.
Return a new RDD that has exactly numPartitions partitions.
Return a new RDD that has exactly numPartitions partitions.
Can increase or decrease the level of parallelism in this RDD. Internally, this uses a shuffle to redistribute data.
If you are decreasing the number of partitions in this RDD, consider using coalesce,
which can avoid performing a shuffle.
Return a sampled subset of this RDD.
Return a sampled subset of this RDD.
can elements be sampled multiple times (replaced when sampled out)
expected size of the sample as a fraction of this RDD's size without replacement: probability that each element is chosen; fraction must be [0, 1] with replacement: expected number of times each element is chosen; fraction must be greater than or equal to 0
seed for the random number generator
This is NOT guaranteed to provide exactly the fraction of the count of the given RDD.
Save this RDD as a SequenceFile of serialized objects.
Save this RDD as a compressed text file, using string representations of elements.
Save this RDD as a text file, using string representations of elements.
Assign a name to this RDD
Return this RDD sorted by the given key function.
The SparkContext that created this RDD.
Return an RDD with the elements from this that are not in other.
Return an RDD with the elements from this that are not in other.
Return an RDD with the elements from this that are not in other.
Return an RDD with the elements from this that are not in other.
Uses this partitioner/partition size, because even if other is huge, the resulting
RDD will be <= us.
Take the first num elements of the RDD.
Take the first num elements of the RDD. It works by first scanning one partition, and use the results from that partition to estimate the number of additional partitions needed to satisfy the limit.
Due to complications in the internal implementation, this method will raise
an exception if called on an RDD of Nothing or Null.
This method should only be used if the resulting array is expected to be small, as all the data is loaded into the driver's memory.
Returns the first k (smallest) elements from this RDD as defined by the specified implicit Ordering[T] and maintains the ordering.
Returns the first k (smallest) elements from this RDD as defined by the specified implicit Ordering[T] and maintains the ordering. This does the opposite of top. For example:
sc.parallelize(Seq(10, 4, 2, 12, 3)).takeOrdered(1) // returns Array(2) sc.parallelize(Seq(2, 3, 4, 5, 6)).takeOrdered(2) // returns Array(2, 3)
k, the number of elements to return
the implicit ordering for T
an array of top elements
This method should only be used if the resulting array is expected to be small, as all the data is loaded into the driver's memory.
Return a fixed-size sampled subset of this RDD in an array
Return a fixed-size sampled subset of this RDD in an array
whether sampling is done with replacement
size of the returned sample
seed for the random number generator
sample of specified size in an array
this method should only be used if the resulting array is expected to be small, as all the data is loaded into the driver's memory.
A description of this RDD and its recursive dependencies for debugging.
Return an iterator that contains all of the elements in this RDD.
Return an iterator that contains all of the elements in this RDD.
The iterator will consume as much memory as the largest partition in this RDD.
This results in multiple Spark jobs, and if the input RDD is the result of a wide transformation (e.g. join with different partitioners), to avoid recomputing the input RDD should be cached first.
Returns the top k (largest) elements from this RDD as defined by the specified implicit Ordering[T] and maintains the ordering.
Returns the top k (largest) elements from this RDD as defined by the specified implicit Ordering[T] and maintains the ordering. This does the opposite of takeOrdered. For example:
sc.parallelize(Seq(10, 4, 2, 12, 3)).top(1) // returns Array(12) sc.parallelize(Seq(2, 3, 4, 5, 6)).top(2) // returns Array(6, 5)
k, the number of top elements to return
the implicit ordering for T
an array of top elements
This method should only be used if the resulting array is expected to be small, as all the data is loaded into the driver's memory.
Aggregates the elements of this RDD in a multi-level tree pattern.
Aggregates the elements of this RDD in a multi-level tree pattern.
suggested depth of the tree (default: 2)
Reduces the elements of this RDD in a multi-level tree pattern.
Reduces the elements of this RDD in a multi-level tree pattern.
suggested depth of the tree (default: 2)
Return the union of this RDD and another one.
Return the union of this RDD and another one. Any identical elements will appear multiple
times (use .distinct() to eliminate them).
Mark the RDD as non-persistent, and remove all blocks for it from memory and disk.
Mark the RDD as non-persistent, and remove all blocks for it from memory and disk.
Whether to block until all blocks are deleted.
This RDD.
Zips this RDD with another one, returning key-value pairs with the first element in each RDD, second element in each RDD, etc.
Zips this RDD with another one, returning key-value pairs with the first element in each RDD, second element in each RDD, etc. Assumes that the two RDDs have the *same number of partitions* and the *same number of elements in each partition* (e.g. one was made through a map on the other).
Zip this RDD's partitions with one (or more) RDD(s) and return a new RDD by applying a function to the zipped partitions.
Zip this RDD's partitions with one (or more) RDD(s) and return a new RDD by applying a function to the zipped partitions. Assumes that all the RDDs have the *same number of partitions*, but does *not* require them to have the same number of elements in each partition.
Zips this RDD with its element indices.
Zips this RDD with its element indices. The ordering is first based on the partition index and then the ordering of items within each partition. So the first item in the first partition gets index 0, and the last item in the last partition receives the largest index.
This is similar to Scala's zipWithIndex but it uses Long instead of Int as the index type. This method needs to trigger a spark job when this RDD contains more than one partitions.
Some RDDs, such as those returned by groupBy(), do not guarantee order of elements in a partition. The index assigned to each element is therefore not guaranteed, and may even change if the RDD is reevaluated. If a fixed ordering is required to guarantee the same index assignments, you should sort the RDD with sortByKey() or save it to a file.
Zips this RDD with generated unique Long ids.
Zips this RDD with generated unique Long ids. Items in the kth partition will get ids k, n+k, 2*n+k, ..., where n is the number of partitions. So there may exist gaps, but this method won't trigger a spark job, which is different from org.apache.spark.rdd.RDD#zipWithIndex.
Some RDDs, such as those returned by groupBy(), do not guarantee order of elements in a partition. The unique ID assigned to each element is therefore not guaranteed, and may even change if the RDD is reevaluated. If a fixed ordering is required to guarantee the same index assignments, you should sort the RDD with sortByKey() or save it to a file.
A Resilient Distributed Dataset (RDD), the basic abstraction in Spark. Represents an immutable, partitioned collection of elements that can be operated on in parallel. This class contains the basic operations available on all RDDs, such as
map,filter, andpersist. In addition, org.apache.spark.rdd.PairRDDFunctions contains operations available only on RDDs of key-value pairs, such asgroupByKeyandjoin; org.apache.spark.rdd.DoubleRDDFunctions contains operations available only on RDDs of Doubles; and org.apache.spark.rdd.SequenceFileRDDFunctions contains operations available on RDDs that can be saved as SequenceFiles. All operations are automatically available on any RDD of the right type (e.g. RDD[(Int, Int)] through implicit.Internally, each RDD is characterized by five main properties:
All of the scheduling and execution in Spark is done based on these methods, allowing each RDD to implement its own way of computing itself. Indeed, users can implement custom RDDs (e.g. for reading data from a new storage system) by overriding these functions. Please refer to the Spark paper for more details on RDD internals.