spark教程(二)-python基础编程

>>> spark.read.text('README.md')
DataFrame[value: string]

# dataframe 的属性
['__class__', '__delattr__', '__dict__', '__doc__', '__format__', '__getattr__', '__getattribute__', '__getitem__', '__hash__', '__init__', '__module__', '__new__', '__reduce__', '__reduce_
ex__', '__repr__', '__setattr__', '__sizeof__', '__str__', '__subclasshook__', '__weakref__', '_collectAsArrow', '_jcols', '_jdf', '_jmap', '_jseq', '_lazy_rdd', '_repr_html_', '_sc', '_schema', 
'_sort_cols', '_support_repr_html', 'agg', 'alias', 'approxQuantile', 'cache', 'checkpoint', 'coalesce', 'colRegex', 'collect', 'columns', 'corr', 'count', 'cov', 'createGlobalTempView', 
'createOrReplaceGlobalTempView', 'createOrReplaceTempView', 'createTempView', 'crossJoin', 'crosstab', 'cube', 'describe', 'distinct', 'drop', 'dropDuplicates', 'drop_duplicates', 'dropna', 
'dtypes', 'exceptAll', 'explain', 'fillna', 'filter', 'first', 'foreach', 'foreachPartition', 'freqItems', 'groupBy', 'groupby', 'head', 'hint', 'intersect', 'intersectAll', 'isLocal', 
'isStreaming', 'is_cached', 'join', 'limit', 'localCheckpoint', 'na', 'orderBy', 'persist', 'printSchema', 'randomSplit', 'rdd', 'registerTempTable', 'repartition', 'repartitionByRange',
'replace', 'rollup', 'sample', 'sampleBy', 'schema', 'select', 'selectExpr', 'show', 'sort', 'sortWithinPartitions', 'sql_ctx', 'stat', 'storageLevel', 'subtract', 'summary', 'take', 'toDF', 
'toJSON', 'toLocalIterator', 'toPandas', 'union', 'unionAll', 'unionByName', 'unpersist', 'where', 'withColumn', 'withColumnRenamed', 'withWatermark', 'write', 'writeStream']

>>> dir(spark)
['Builder', '__class__', '__delattr__', '__dict__', '__doc__', '__enter__', '__exit__', '__format__', '__getattribute__', '__hash__', '__init__', '__module__', '__new__', '__reduce__', '__r
educe_ex__', '__repr__', '__setattr__', '__sizeof__', '__str__', '__subclasshook__', '__weakref__', '_convert_from_pandas', '_createFromLocal', '_createFromRDD', '_create_from_pandas_with_arrow', 
'_create_shell_session', '_get_numpy_record_dtype', '_inferSchema', '_inferSchemaFromList', '_instantiatedSession', '_jsc', '_jsparkSession', '_jvm', '_jwrapped', '_repr_html_', '_sc', '_wrapped', 
'builder', 'catalog', 'conf', 'createDataFrame', 'newSession', 'range', 'read', 'readStream', 'sparkContext', 'sql', 'stop', 'streams', 'table', 'udf', 'version']

spark.sparkContext  # 即 sc

>>> sc.textFile('README.md').count()
>>> spark.read.text('README.md').count()

[root@hadoop10 spark]# bin/pyspark --help
Usage: ./bin/pyspark [options]

Options:
  --master MASTER_URL         spark://host:port, mesos://host:port, yarn,               # 设定 master，即在哪里运行 spark，
                                                                                        # mesos://host:port一般不用；yarn需要把spark部署到yarn上
                              k8s://https://host:port, or local (Default: local[*]).    # local 本地模式，local 表示单线程，local[num]表示num个进程，
                                                                                        # local[*]表示服务器cpu是几核就是几个进程
  --deploy-mode DEPLOY_MODE   Whether to launch the driver program locally ("client") or
                              on one of the worker machines inside the cluster ("cluster")
                              (Default: client).
  --class CLASS_NAME          Your application's main class (for Java / Scala apps).    # 要执行的 class 类名
  --name NAME                 A name of your application.
  --jars JARS                 Comma-separated list of jars to include on the driver
                              and executor classpaths.
  --packages                  Comma-separated list of maven coordinates of jars to include　　# 逗号隔开的 maven 列表，给 当前会话 添加依赖
                              on the driver and executor classpaths. Will search the local
                              maven repo, then maven central and any additional remote
                              repositories given by --repositories. The format for the
                              coordinates should be groupId:artifactId:version.
  --exclude-packages          Comma-separated list of groupId:artifactId, to exclude while
                              resolving the dependencies provided in --packages to avoid
                              dependency conflicts.
  --repositories              Comma-separated list of additional remote repositories to
                              search for the maven coordinates given with --packages.
  --py-files PY_FILES         Comma-separated list of .zip, .egg, or .py files to place　　# 逗号隔开的 文件列表，替代 PYTHONPATH 的作用，也就是说如果不设置 PYTHONPATH，就需要这个参数，才能导入 文件中的模块
                              on the PYTHONPATH for Python apps.
  --files FILES               Comma-separated list of files to be placed in the working
                              directory of each executor. File paths of these files
                              in executors can be accessed via SparkFiles.get(fileName).

  --conf PROP=VALUE           Arbitrary Spark configuration property.
  --properties-file FILE      Path to a file from which to load extra properties. If not
                              specified, this will look for conf/spark-defaults.conf.

  --driver-memory MEM         Memory for driver (e.g. 1000M, 2G) (Default: 1024M).
  --driver-java-options       Extra Java options to pass to the driver.
  --driver-library-path       Extra library path entries to pass to the driver.
  --driver-class-path         Extra class path entries to pass to the driver. Note that
                              jars added with --jars are automatically included in the
                              classpath.

  --executor-memory MEM       Memory per executor (e.g. 1000M, 2G) (Default: 1G).

  --proxy-user NAME           User to impersonate when submitting the application.
                              This argument does not work with --principal / --keytab.

  --help, -h                  Show this help message and exit.
  --verbose, -v               Print additional debug output.
  --version,                  Print the version of current Spark.

 Cluster deploy mode only:
  --driver-cores NUM          Number of cores used by the driver, only in cluster mode
                              (Default: 1).

 Spark standalone or Mesos with cluster deploy mode only:
  --supervise                 If given, restarts the driver on failure.
  --kill SUBMISSION_ID        If given, kills the driver specified.
  --status SUBMISSION_ID      If given, requests the status of the driver specified.

 Spark standalone and Mesos only:
  --total-executor-cores NUM  Total cores for all executors.

 Spark standalone and YARN only:
  --executor-cores NUM        Number of cores per executor. (Default: 1 in YARN mode,
                              or all available cores on the worker in standalone mode)

 YARN-only:
  --queue QUEUE_NAME          The YARN queue to submit to (Default: "default").
  --num-executors NUM         Number of executors to launch (Default: 2).
                              If dynamic allocation is enabled, the initial number of
                              executors will be at least NUM.
  --archives ARCHIVES         Comma separated list of archives to be extracted into the
                              working directory of each executor.
  --principal PRINCIPAL       Principal to be used to login to KDC, while running on
                              secure HDFS.
  --keytab KEYTAB             The full path to the file that contains the keytab for the
                              principal specified above. This keytab will be copied to
                              the node running the Application Master via the Secure
                              Distributed Cache, for renewing the login tickets and the
                              delegation tokens periodically.

[root@hadoop10 spark]# bin/pyspark 
Python 2.7.12 (default, Oct  2 2019, 19:43:15) 
[GCC 4.4.7 20120313 (Red Hat 4.4.7-4)] on linux2
Type "help", "copyright", "credits" or "license" for more information.
19/10/09 18:10:53 WARN NativeCodeLoader: Unable to load native-hadoop library for your platform... using builtin-java classes where applicable
Using Spark's default log4j profile: org/apache/spark/log4j-defaults.properties
Setting default log level to "WARN".
To adjust logging level use sc.setLogLevel(newLevel). For SparkR, use setLogLevel(newLevel).
Welcome to
      ____              __
     / __/__  ___ _____/ /__
    _\ \/ _ \/ _ `/ __/  '_/
   /__ / .__/\_,_/_/ /_/\_\   version 2.4.4
      /_/

Using Python version 2.7.12 (default, Oct  2 2019 19:43:15)
SparkSession available as 'spark'.  # 自带 spark

>>> distFile = sc.textFile('README.md')
>>> distFile.map(lambda x: len(x)).reduce(lambda a, b: a + b)
3847                                                                            
>>> distFile.count()
105

class SparkContext(__builtin__.object):
    def __init__(self, master=None, appName=None, sparkHome=None, pyFiles=None, environment=None, batchSize=0, serializer=PickleSerializer(), conf=None, gateway=None, jsc=None, profiler_cls=<cl
    ass 'pyspark.profiler.BasicProfiler'>)
    '''Create a new SparkContext. At least the master and app name should be set,
     |      either through the named parameters here or through C{conf}.
     |
     |      :param master: Cluster URL to connect to
     |             (e.g. mesos://host:port, spark://host:port, local[4]).   local 表示本地运行，4 表示使用4个 cpu核
     |      :param appName: A name for your job, to display on the cluster web UI.
     |      :param sparkHome: Location where Spark is installed on cluster nodes.
     |      :param pyFiles: Collection of .zip or .py files to send to the cluster
     |             and add to PYTHONPATH.  These can be paths on the local file
     |             system or HDFS, HTTP, HTTPS, or FTP URLs.
     |      :param environment: A dictionary of environment variables to set on
     |             worker nodes.
     |      :param batchSize: The number of Python objects represented as a single
     |             Java object. Set 1 to disable batching, 0 to automatically choose
     |             the batch size based on object sizes, or -1 to use an unlimited
     |             batch size
     |      :param serializer: The serializer for RDDs.
     |      :param conf: A L{SparkConf} object setting Spark properties.
     |      :param gateway: Use an existing gateway and JVM, otherwise a new JVM
     |             will be instantiated.
     |      :param jsc: The JavaSparkContext instance (optional).
     |      :param profiler_cls: A class of custom Profiler used to do profiling
     |             (default is pyspark.profiler.BasicProfiler).
     |
     |
     |      >>> from pyspark.context import SparkContext
     |      >>> sc = SparkContext('local', 'test')
     |
     |      >>> sc2 = SparkContext('local', 'test2') # doctest: +IGNORE_EXCEPTION_DETAIL
     |      Traceback (most recent call last):
     |          ...
     |      ValueError:...'''

from pyspark import SparkContext, SparkConf
conf = SparkConf().setAppName('myapp1').setMaster('local[4]')   # 设定 appname 和 master
sc = SparkContext(conf=conf)

## 或者直接这样
sc = SparkContext("spark://hadoop10:7077")

class SparkSession(__builtin__.object):
     def __init__(self, sparkContext, jsparkSession=None):
     ''' Creates a new SparkSession.
     |
     |      >>> from datetime import datetime
     |      >>> spark = SparkSession(sc)
     |      >>> allTypes = sc.parallelize([Row(i=1, s="string", d=1.0, l=1,
     |      ...     b=True, list=[1, 2, 3], dict={"s": 0}, row=Row(a=1),
     |      ...     time=datetime(2014, 8, 1, 14, 1, 5))])
     |      >>> df = allTypes.toDF()
     |      >>> df.createOrReplaceTempView("allTypes")
     |      >>> spark.sql('select i+1, d+1, not b, list[1], dict["s"], time, row.a '
     |      ...            'from allTypes where b and i > 0').collect()
     |      [Row((i + CAST(1 AS BIGINT))=2, (d + CAST(1 AS DOUBLE))=2.0, (NOT b)=False, list[1]=2,             dict[s]=0, time=datetime.datetime(2014, 8, 1, 14, 1, 5), a=1)]
     |      >>> df.rdd.map(lambda x: (x.i, x.s, x.d, x.l, x.b, x.time, x.row.a, x.list)).collect()
     |      [(1, u'string', 1.0, 1, True, datetime.datetime(2014, 8, 1, 14, 1, 5), 1, [1, 2, 3])]'''

from pyspark.sql import SparkSession

conf = SparkConf().setAppName('myapp1').setMaster('local[4]')   # 设定 appname 和 master
sess = SparkSession.builder.config(conf=conf)

## 或者这样
from pyspark import SparkContext, SparkConf
conf = SparkConf().setAppName('myapp1').setMaster('local[4]')   # 设定 appname 和 master
sc = SparkContext(conf=conf)
sess = SparkSession(sc)

data = range(10000)
distData = sc.parallelize(data)
distData.reduce(lambda a, b: a+b)

distData2 = sc.parallelize(data, 100)   # 切 100 分

textFile(name, minPartitions=None, use_unicode=True)[source]
Read a text file from HDFS, a local file system (available on all nodes), or any Hadoop-supported file system URI, and return it as an RDD of Strings.

If use_unicode is False, the strings will be kept as str (encoding as utf-8), which is faster and smaller than unicode. (Added in Spark 1.2)

distFile = sc.textFile('README.md')
distFile = sc.textFile('xx.csv')
distFile = sc.textFile('logs')      # 支持文件夹
# textFile("/my/directory/*.txt")   # 支持通配符
# textFile("/my/directory/*.gz")    # 支持压缩文件

type(sc.textFile('README.md'))      # <class 'pyspark.rdd.RDD'>


distFile.map(lambda x: int(x[-1])).reduce(lambda a, b: a + b)
distFile.map(lambda x: len(x)).reduce(lambda a, b: a + b)   # map 的输入是每一行，map 的作用是求每一行的 len
                                                            # reduce 的输入是两个数字，reduce 的作用是求这两个数的和，
                                                            # 也就是把 所有行的 len 逐次求当前元素（当前累计 len 的和）与下一元素的和

distFile = sc.wholeTextFiles('files')   # 读取 files 文件夹下所有文件内容，返回 (filename， filecontent) 键值对

distFile = sc.textFile('README.md')
m = distFile.map(lambda x: len(x))      # map 是 转换 操作，并不立即执行
m.cache()       # 把 map 的输出缓存到内存中，其实 cache 就是 执行 操作

rdd = sc.parallelize([2, 3, 4])
rdd.map(lambda x: x + 1).collect()          # [3, 4, 5]

rdd = sc.parallelize([2, 3, 4])
rdd.map(lambda x: x > 3).collect()          # [False, False, True]

rdd = sc.parallelize([2, 3, 4])
rdd.flatMap(lambda x: range(1, x)).collect()            # [1, 1, 2, 1, 2, 3]
# range(1, 2): 1
# range(1, 3): 1, 2
# range(1, 4): 1, 2, 3

### vs map
rdd.map(lambda x: range(1, x)).collect()                # [[1], [1, 2], [1, 2, 3]]

rdd = sc.parallelize([1,2,3,4,5], 3)    # 分 3 个区
def f(iterator): yield sum(iterator)    # 必须是生成器，即 yield，不能 return
rdd.mapPartitions(f).collect()          # [1, 5, 9]

rdd = sc.parallelize(range(15), 13)    # 分 13 个区
def f(splitIndex, iterator): yield splitIndex
rdd.mapPartitionsWithIndex(f).collect() # [0, 1, 2, 3, 4, 5, 6, 7, 8, 9, 10, 11, 12]

x = sc.parallelize([1,2,3])
y = sc.parallelize([4,5])
x.cartesian(y).collect()        # [(1, 4), (1, 5), (2, 4), (2, 5), (3, 4), (3, 5)]

rdd = sc.parallelize([("a", 1), ("b", 1), ("a", 1)])
def f(x, y): return x + y
rdd.reduceByKey(f).collect()        # [('a', 2), ('b', 1)]

tmp = [('a', 4), ('b', 3), ('c', 2), ('D', 1)]
sc.parallelize(tmp).sortByKey(True, 1).collect()    # 升序[('D', 1), ('a', 4), ('b', 3), ('c', 2)] 1代表分区数
sc.parallelize(tmp).sortByKey(True, 2, keyfunc=lambda k:k.lower()).collect()   # 升序[('a', 4), ('b', 3), ('c', 2), ('D', 1)] D跑到后面了

sc.parallelize(tmp).sortByKey(False, 2, keyfunc=lambda k:k.lower()).collect()# 降序[('D', 1), ('c', 2), ('b', 3), ('a', 4)]

x = sc.parallelize([("a", 1), ("b", 4)])
y = sc.parallelize([("a", 2), ("a", 3)])
x.join(y).collect()     # [('a', (1, 2)), ('a', (1, 3))]

sc.parallelize([9,7,3,2,6,4]).takeOrdered(3)    # [2, 3, 4]
sc.parallelize([9,7,3,2,6,4]).takeOrdered(3, key=lambda x: -x)   # [9, 7, 6]
## 过程如下
#  9,  7,  3,  2,  6,  4  ## 原数据
# -9, -7, -3, -2, -6, -4  ## 经过 keyfunc 处理后的数据
# -9, -7, -6, -4, -3, -2  ## 对处理后的数据升序排序
# -9, -7, -6              ## 取前3个
#  9,  7,  6              ## 对应到原数据

sc.parallelize([1, 2, 3, 4, 5]).foreach(print)  # 并行打印，不按顺序输出
# 1
# 2
# 4
# 5
# 3

sc.parallelize([1,2,3,4,5]).reduce(lambda x, y: x + y)  # 15 求和

x = sc.parallelize([1,2,3])
neutral_zero_value = 0  # 0 for sum, 1 for multiplication
y = x.fold(neutral_zero_value, lambda obj, accumulated: accumulated + obj) # computes cumulative sum
print(x.collect())  # [1,2,3]
print(y)            # 6

sc.parallelize([1,2,3,1,2,5,3,2,3,2]).countByValue().items()    # [(1, 2), (2, 4), (3, 3), (5, 1)]

# 输入 k-v 不按 value 统计，按 k-v 统计
sc.parallelize([('a', 1), ('b', 1)]).countByValue().items()     # [(('a', 1), 1), (('b', 1), 1)]

counts.saveAsTextFile('/usr/lib/spark/out')

sc.parallelize([("a",1), ("b",1), ("a", 3)]).countByKey()   # defaultdict(<type 'int'>, {'a': 2, 'b': 1})

dictdata= sc.parallelize([("a",1), ("b",1), ("a", 3)]).countByKey()
dictdata.items()        # [('a', 2), ('b', 1)]

# python can call pyspark directly
export PYTHONPATH=$SPARK_HOME/python:$SPARK_HOME/python/pyspark:$SPARK_HOME/python/lib/py4j-0.10.4-src.zip:$PYTHONPATH

from __future__ import print_function
from pyspark import *
import os
print(os.environ['SPARK_HOME'])
print(os.environ['HADOOP_HOME'])
if __name__ == '__main__':
    sc = SparkContext("spark://hadoop10:7077")
    rdd = sc.parallelize("hello Pyspark world".split(' '))
    counts = rdd.map(lambda word: (word, 1)) \
        .reduceByKey(lambda a, b: a + b)
    counts.saveAsTextFile('/usr/lib/spark/out')
    counts.foreach(print)

    sc.stop()

bin/spark-submit test1.py

python test1.py