github.com/apache/beam/sdks/v2@v2.48.2/python/apache_beam/examples/snippets/transforms/aggregation/combinevalues.py (about)

     1  # coding=utf-8
     2  #
     3  # Licensed to the Apache Software Foundation (ASF) under one or more
     4  # contributor license agreements.  See the NOTICE file distributed with
     5  # this work for additional information regarding copyright ownership.
     6  # The ASF licenses this file to You under the Apache License, Version 2.0
     7  # (the "License"); you may not use this file except in compliance with
     8  # the License.  You may obtain a copy of the License at
     9  #
    10  #    http://www.apache.org/licenses/LICENSE-2.0
    11  #
    12  # Unless required by applicable law or agreed to in writing, software
    13  # distributed under the License is distributed on an "AS IS" BASIS,
    14  # WITHOUT WARRANTIES OR CONDITIONS OF ANY KIND, either express or implied.
    15  # See the License for the specific language governing permissions and
    16  # limitations under the License.
    17  #
    18  
    19  # pytype: skip-file
    20  
    21  
    22  def combinevalues_simple(test=None):
    23    # [START combinevalues_simple]
    24    import apache_beam as beam
    25  
    26    with beam.Pipeline() as pipeline:
    27      total = (
    28          pipeline
    29          | 'Create produce counts' >> beam.Create([
    30              ('🥕', [3, 2]),
    31              ('🍆', [1]),
    32              ('🍅', [4, 5, 3]),
    33          ])
    34          | 'Sum' >> beam.CombineValues(sum)
    35          | beam.Map(print))
    36      # [END combinevalues_simple]
    37      if test:
    38        test(total)
    39  
    40  
    41  def combinevalues_function(test=None):
    42    # [START combinevalues_function]
    43    import apache_beam as beam
    44  
    45    def saturated_sum(values):
    46      max_value = 8
    47      return min(sum(values), max_value)
    48  
    49    with beam.Pipeline() as pipeline:
    50      saturated_total = (
    51          pipeline
    52          | 'Create plant counts' >> beam.Create([
    53              ('🥕', [3, 2]),
    54              ('🍆', [1]),
    55              ('🍅', [4, 5, 3]),
    56          ])
    57          | 'Saturated sum' >> beam.CombineValues(saturated_sum)
    58          | beam.Map(print))
    59      # [END combinevalues_function]
    60      if test:
    61        test(saturated_total)
    62  
    63  
    64  def combinevalues_lambda(test=None):
    65    # [START combinevalues_lambda]
    66    import apache_beam as beam
    67  
    68    with beam.Pipeline() as pipeline:
    69      saturated_total = (
    70          pipeline
    71          | 'Create plant counts' >> beam.Create([
    72              ('🥕', [3, 2]),
    73              ('🍆', [1]),
    74              ('🍅', [4, 5, 3]),
    75          ])
    76          | 'Saturated sum' >>
    77          beam.CombineValues(lambda values: min(sum(values), 8))
    78          | beam.Map(print))
    79      # [END combinevalues_lambda]
    80      if test:
    81        test(saturated_total)
    82  
    83  
    84  def combinevalues_multiple_arguments(test=None):
    85    # [START combinevalues_multiple_arguments]
    86    import apache_beam as beam
    87  
    88    with beam.Pipeline() as pipeline:
    89      saturated_total = (
    90          pipeline
    91          | 'Create plant counts' >> beam.Create([
    92              ('🥕', [3, 2]),
    93              ('🍆', [1]),
    94              ('🍅', [4, 5, 3]),
    95          ])
    96          | 'Saturated sum' >> beam.CombineValues(
    97              lambda values, max_value: min(sum(values), max_value), max_value=8)
    98          | beam.Map(print))
    99      # [END combinevalues_multiple_arguments]
   100      if test:
   101        test(saturated_total)
   102  
   103  
   104  def combinevalues_side_inputs_singleton(test=None):
   105    # [START combinevalues_side_inputs_singleton]
   106    import apache_beam as beam
   107  
   108    with beam.Pipeline() as pipeline:
   109      max_value = pipeline | 'Create max_value' >> beam.Create([8])
   110  
   111      saturated_total = (
   112          pipeline
   113          | 'Create plant counts' >> beam.Create([
   114              ('🥕', [3, 2]),
   115              ('🍆', [1]),
   116              ('🍅', [4, 5, 3]),
   117          ])
   118          | 'Saturated sum' >> beam.CombineValues(
   119              lambda values,
   120              max_value: min(sum(values), max_value),
   121              max_value=beam.pvalue.AsSingleton(max_value))
   122          | beam.Map(print))
   123      # [END combinevalues_side_inputs_singleton]
   124      if test:
   125        test(saturated_total)
   126  
   127  
   128  def combinevalues_side_inputs_iter(test=None):
   129    # [START combinevalues_side_inputs_iter]
   130    import apache_beam as beam
   131  
   132    def bounded_sum(values, data_range):
   133      min_value = min(data_range)
   134      result = sum(values)
   135      if result < min_value:
   136        return min_value
   137      max_value = max(data_range)
   138      if result > max_value:
   139        return max_value
   140      return result
   141  
   142    with beam.Pipeline() as pipeline:
   143      data_range = pipeline | 'Create data_range' >> beam.Create([2, 4, 8])
   144  
   145      bounded_total = (
   146          pipeline
   147          | 'Create plant counts' >> beam.Create([
   148              ('🥕', [3, 2]),
   149              ('🍆', [1]),
   150              ('🍅', [4, 5, 3]),
   151          ])
   152          | 'Bounded sum' >> beam.CombineValues(
   153              bounded_sum, data_range=beam.pvalue.AsIter(data_range))
   154          | beam.Map(print))
   155      # [END combinevalues_side_inputs_iter]
   156      if test:
   157        test(bounded_total)
   158  
   159  
   160  def combinevalues_side_inputs_dict(test=None):
   161    # [START combinevalues_side_inputs_dict]
   162    import apache_beam as beam
   163  
   164    def bounded_sum(values, data_range):
   165      min_value = data_range['min']
   166      result = sum(values)
   167      if result < min_value:
   168        return min_value
   169      max_value = data_range['max']
   170      if result > max_value:
   171        return max_value
   172      return result
   173  
   174    with beam.Pipeline() as pipeline:
   175      data_range = pipeline | 'Create data_range' >> beam.Create([
   176          ('min', 2),
   177          ('max', 8),
   178      ])
   179  
   180      bounded_total = (
   181          pipeline
   182          | 'Create plant counts' >> beam.Create([
   183              ('🥕', [3, 2]),
   184              ('🍆', [1]),
   185              ('🍅', [4, 5, 3]),
   186          ])
   187          | 'Bounded sum' >> beam.CombineValues(
   188              bounded_sum, data_range=beam.pvalue.AsDict(data_range))
   189          | beam.Map(print))
   190      # [END combinevalues_side_inputs_dict]
   191      if test:
   192        test(bounded_total)
   193  
   194  
   195  def combinevalues_combinefn(test=None):
   196    # [START combinevalues_combinefn]
   197    import apache_beam as beam
   198  
   199    class AverageFn(beam.CombineFn):
   200      def create_accumulator(self):
   201        return {}
   202  
   203      def add_input(self, accumulator, input):
   204        # accumulator == {}
   205        # input == '🥕'
   206        if input not in accumulator:
   207          accumulator[input] = 0  # {'🥕': 0}
   208        accumulator[input] += 1  # {'🥕': 1}
   209        return accumulator
   210  
   211      def merge_accumulators(self, accumulators):
   212        # accumulators == [
   213        #     {'🥕': 1, '🍅': 1},
   214        #     {'🥕': 1, '🍅': 1, '🍆': 1},
   215        # ]
   216        merged = {}
   217        for accum in accumulators:
   218          for item, count in accum.items():
   219            if item not in merged:
   220              merged[item] = 0
   221            merged[item] += count
   222        # merged == {'🥕': 2, '🍅': 2, '🍆': 1}
   223        return merged
   224  
   225      def extract_output(self, accumulator):
   226        # accumulator == {'🥕': 2, '🍅': 2, '🍆': 1}
   227        total = sum(accumulator.values())  # 5
   228        percentages = {item: count / total for item, count in accumulator.items()}
   229        # percentages == {'🥕': 0.4, '🍅': 0.4, '🍆': 0.2}
   230        return percentages
   231  
   232    with beam.Pipeline() as pipeline:
   233      percentages_per_season = (
   234          pipeline
   235          | 'Create produce' >> beam.Create([
   236              ('spring', ['🥕', '🍅', '🥕', '🍅', '🍆']),
   237              ('summer', ['🥕', '🍅', '🌽', '🍅', '🍅']),
   238              ('fall', ['🥕', '🥕', '🍅', '🍅']),
   239              ('winter', ['🍆', '🍆']),
   240          ])
   241          | 'Average' >> beam.CombineValues(AverageFn())
   242          | beam.Map(print))
   243      # [END combinevalues_combinefn]
   244      if test:
   245        test(percentages_per_season)