Tablestore:Aggregation

Parameters

Example

Calculate the minimum score among people who are 18 years old.

query = TermQuery('age', 18)
agg = Min('score', name='min')

search_response = client.search(table_name, index_name,
    SearchQuery(query, limit=0, aggs=[agg]),
    ColumnsToGet(return_type=ColumnReturnType.ALL_FROM_INDEX))

for agg_result in search_response.agg_results:
    print("name: %s, value: %s" % (agg_result.name, str(agg_result.value)))

Parameters

Example

Calculate the maximum score among people who are 18 years old. If a person does not have a score, a default value of 0 is used.

query = TermQuery('age', 18)
agg = Max('score', missing_value=0, name='max')

search_response = client.search(table_name, index_name,
    SearchQuery(query, limit=0, aggs=[agg]),
    ColumnsToGet(return_type=ColumnReturnType.ALL_FROM_INDEX))

for agg_result in search_response.agg_results:
    print("name: %s, value: %s" % (agg_result.name, str(agg_result.value)))

Parameters

Example

Calculate the sum of all scores for people who are 18 years old.

query = TermQuery('age', 18)
agg = Sum('score', name='sum')

search_response = client.search(table_name, index_name,
    SearchQuery(query, limit=0, aggs=[agg]),
    ColumnsToGet(return_type=ColumnReturnType.ALL_FROM_INDEX))

for agg_result in search_response.agg_results:
    print("name: %s, value: %s" % (agg_result.name, str(agg_result.value)))

Parameters

Example

Calculate the average score for people who are 18 years old.

query = TermQuery('age', 18)
agg = Avg('score', name='avg')

search_response = client.search(table_name, index_name,
    SearchQuery(query, limit=0, aggs=[agg]),
    ColumnsToGet(return_type=ColumnReturnType.ALL_FROM_INDEX))

for agg_result in search_response.agg_results:
    print("name: %s, value: %s" % (agg_result.name, str(agg_result.value)))

Parameters

Example

Count the number of people who are 18 years old and have a recorded exam score.

query = TermQuery('age', 18)
agg = Count('score', name='count')

search_response = client.search(table_name, index_name,
    SearchQuery(query, limit=0, aggs=[agg]),
    ColumnsToGet(return_type=ColumnReturnType.ALL_FROM_INDEX))

for agg_result in search_response.agg_results:
    print("name: %s, value: %s" % (agg_result.name, str(agg_result.value)))

Parameters

Example

Count the number of distinct names among people who are 18 years old.

query = TermQuery('age', 18)
agg = DistinctCount('name', name='distinct_name')

search_response = client.search(table_name, index_name,
    SearchQuery(query, limit=0, aggs=[agg]),
    ColumnsToGet(return_type=ColumnReturnType.ALL_FROM_INDEX))

for agg_result in search_response.agg_results:
    print("name: %s, value: %s" % (agg_result.name, str(agg_result.value)))

Parameters

Example

query = TermQuery('product', '10010')
agg = Percentiles('latency', percentiles_list=[50, 90, 95])

search_response = client.search(table_name, index_name,
    SearchQuery(query, limit=0, aggs=[agg]),
    ColumnsToGet(return_type=ColumnReturnType.ALL_FROM_INDEX))

for agg_result in search_response.agg_results:
    print("name: %s" % agg_result.name)
    for item in agg_result.value:
        print("  percentile: %s, value: %s" % (str(item.key), str(item.value)))

Parameters

Example 1

Group people who are 18 years old by score, and retrieve the 10 most common scores and the number of people for each score.

query = TermQuery('age', 18)
group_by = GroupByField('score', size=10)

search_response = client.search(table_name, index_name,
    SearchQuery(query, limit=0, group_bys=[group_by]),
    ColumnsToGet(return_type=ColumnReturnType.ALL_FROM_INDEX))

for group_by_result in search_response.group_by_results:
    print("name: %s" % group_by_result.name)
    for item in group_by_result.items:
        print("  key: %s, count: %d" % (item.key, item.row_count))

Example 2

Group people who are 18 years old by score, and retrieve the two least common scores and the number of people for each score.

group_by = GroupByField('score', size=2, group_by_sort=[RowCountSort(sort_order=SortOrder.ASC)])

search_response = client.search(table_name, index_name,
    SearchQuery(TermQuery('age', 18), limit=0, get_total_count=True, group_bys=[group_by]),
    ColumnsToGet(return_type=ColumnReturnType.ALL_FROM_INDEX))

for group_by_result in search_response.group_by_results:
    print("name: %s" % group_by_result.name)
    for item in group_by_result.items:
        print("  key: %s, count: %d" % (item.key, item.row_count))

Example 3

Group people who are 18 years old by score, retrieve the two most common scores and their counts, and then get information about the top three individuals in each group, sorted by primary key.

sort = RowCountSort(sort_order=SortOrder.DESC)
sub_agg = [TopRows(limit=3, sort=Sort([PrimaryKeySort(sort_order=SortOrder.DESC)]), name='top_rows')]
group_by = GroupByField('score', size=2, group_by_sort=[sort], sub_aggs=sub_agg)

search_response = client.search(table_name, index_name,
    SearchQuery(TermQuery('age', 18), limit=0, get_total_count=True, group_bys=[group_by]),
    ColumnsToGet(return_type=ColumnReturnType.ALL_FROM_INDEX))

for group_by_result in search_response.group_by_results:
    print("name: %s" % group_by_result.name)
    for item in group_by_result.items:
        print("  key: %s, count: %d" % (item.key, item.row_count))
        for sub_agg in item.sub_aggs:
            print("    sub_agg: %s" % sub_agg.name)
            for entry in sub_agg.value:
                print("      value: %s" % str(entry))

Example 4

Group people who are 18 years old by score and gender.

sort = RowCountSort(sort_order=SortOrder.ASC)
sub_group = GroupByField('sex', size=10, group_by_sort=[sort])
group_by = GroupByField('score', size=10, group_by_sort=[sort], sub_group_bys=[sub_group])

search_response = client.search(table_name, index_name,
    SearchQuery(TermQuery('age', 18), limit=0, get_total_count=True, group_bys=[group_by]),
    ColumnsToGet(return_type=ColumnReturnType.ALL_FROM_INDEX))

for group_by_result in search_response.group_by_results:
    print("name: %s" % group_by_result.name)
    for item in group_by_result.items:
        print("  key: %s, count: %d" % (item.key, item.row_count))
        for sub_group in item.sub_group_bys:
            print("    sub_group: %s" % sub_group.name)
            for sub_item in sub_group.items:
                print("      key: %s, count: %s" % (str(sub_item.key), str(sub_item.row_count)))

Parameters

Example

Count the number of people who are 18 years old and have scores in the [80, 90) and [90, 100) ranges.

query = TermQuery('age', 18)
group_by = GroupByRange(field_name='score', ranges=[(80, 90), (90, 100)])

search_response = client.search(table_name, index_name,
    SearchQuery(query, limit=0, group_bys=[group_by]),
    ColumnsToGet(return_type=ColumnReturnType.ALL_FROM_INDEX))

for group_by_result in search_response.group_by_results:
    print("name: %s" % group_by_result.name)
    for item in group_by_result.items:
        print("  range: %.1f~%.1f, count: %d" % (item.range_from, item.range_to, item.row_count))

Parameters

Example

Count the number of people who are 18 years old and whose homes are within 1 kilometer, and between 1 and 2 kilometers, from their school. The school's coordinates are (31, 116).

query = TermQuery('age', 18)
group_by = GroupByGeoDistance(field_name='address', origin=GeoPoint(31, 116), ranges=[(0, 1000), (1000, 2000)])

search_response = client.search(table_name, index_name,
    SearchQuery(query, limit=0, group_bys=[group_by]),
    ColumnsToGet(return_type=ColumnReturnType.ALL_FROM_INDEX))

for group_by_result in search_response.group_by_results:
    print("name: %s" % group_by_result.name)
    for item in group_by_result.items:
        print("  range: %.1f~%.1f, count: %d" % (item.range_from, item.range_to, item.row_count))

Parameters

Example

Count the number of people who are 18 years old and scored 100 in math, and the number of people who scored 100 in Chinese.

query = TermQuery('age', 18)
filter1 = TermQuery('math', 100)
filter2 = TermQuery('chinese', 100)
filters = [filter1, filter2]
group_by = GroupByFilter(filters)

search_response = client.search(table_name, index_name,
    SearchQuery(query, limit=0, group_bys=[group_by]),
    ColumnsToGet(return_type=ColumnReturnType.ALL_FROM_INDEX))

for group_by_result in search_response.group_by_results:
    print("name: %s" % group_by_result.name)
    i = 0
    for item in group_by_result.items:
        print("  filter: %s=%s, count: %d" % (str(filters[i].field_name), str(filters[i].column_value), item.row_count))
        i += 1

Parameters

Example

query = TermQuery('product', '10010')
group_by = GroupByHistogram(field_name='latency', interval=100, field_range=FieldRange(0, 10000), missing_value=0)

search_response = client.search(table_name, index_name,
    SearchQuery(query, limit=0, group_bys=[group_by]),
    ColumnsToGet(return_type=ColumnReturnType.ALL_FROM_INDEX))

for group_by_result in search_response.group_by_results:
    print("name: %s" % group_by_result.name)
    for item in group_by_result.items:
        print("  key: %s, value: %s" % (item.key, item.value))

Parameters

Example 1

Group people who are 18 years old by the score field, and get the number of rows for each score value.

query = TermQuery('age', 18)
# When a GroupByField is used as a source, you only need to specify the field_name and name parameters. The size parameter cannot be set.
source = GroupByField('score', name='group_by_score')
group_by = GroupByComposite(sources=[source])

search_response = client.search(table_name, index_name,
    SearchQuery(query, limit=0, get_total_count=True, group_bys=[group_by]),
    ColumnsToGet(return_type=ColumnReturnType.NONE))

for group_by_result in search_response.group_by_results:
    print("name: %s" % group_by_result.name)
    print("source_names: %s" % str(group_by_result.source_group_by_names))
    for item in group_by_result.items:
        print("  keys: %s, count: %d" % (str(item.keys), item.row_count))

Example 2

Group by the score and sex fields for individuals aged 18 to obtain the number of rows for each score and sex combination.

query = TermQuery('age', 18)
source1 = GroupByField('score', name='group_by_score')
source2 = GroupByField('sex', name='group_by_sex')
group_by = GroupByComposite(sources=[source1, source2])

search_response = client.search(table_name, index_name,
    SearchQuery(query, limit=0, get_total_count=True, group_bys=[group_by]),
    ColumnsToGet(return_type=ColumnReturnType.NONE))

for group_by_result in search_response.group_by_results:
    print("name: %s" % group_by_result.name)
    print("source_names: %s" % str(group_by_result.source_group_by_names))
    for item in group_by_result.items:
        # The keys parameter is a list of strings that corresponds to the sources in order. If a field value is empty, the corresponding element is None.
        print("  keys: %s, count: %d" % (str(item.keys), item.row_count))

Example 3

Use next_token to retrieve all grouped results in a paginated manner. The results are grouped by the score field, and you can retrieve 2 groups at a time to obtain all results through pagination.

query = TermQuery('age', 18)
source = GroupByField('score')
group_by = GroupByComposite(sources=[source], size=2)

# Initial request
search_response = client.search(table_name, index_name,
    SearchQuery(query, limit=0, get_total_count=True, group_bys=[group_by]),
    ColumnsToGet(return_type=ColumnReturnType.NONE))

group_by_result = search_response.group_by_results[0]
all_items = list(group_by_result.items)

# Paginate to retrieve the remaining results. The operation is complete when next_token is None.
while group_by_result.next_token is not None:
    group_by = GroupByComposite(sources=[source], size=2, next_token=group_by_result.next_token)
    search_response = client.search(table_name, index_name,
        SearchQuery(query, limit=0, get_total_count=True, group_bys=[group_by]),
        ColumnsToGet(return_type=ColumnReturnType.NONE))

    group_by_result = search_response.group_by_results[0]
    all_items.extend(group_by_result.items)

# Print all the grouped results.
for item in all_items:
    print("keys: %s, count: %d" % (str(item.keys), item.row_count))

Example 4

You can use a sub-aggregation to group data by the score field and calculate the maximum value of score for each group.

query = TermQuery('age', 18)
source = GroupByField('score')
sub_agg = Max('score')
group_by = GroupByComposite(sources=[source], sub_aggs=[sub_agg])

search_response = client.search(table_name, index_name,
    SearchQuery(query, limit=0, get_total_count=True, group_bys=[group_by]),
    ColumnsToGet(return_type=ColumnReturnType.NONE))

for group_by_result in search_response.group_by_results:
    print("name: %s" % group_by_result.name)
    for item in group_by_result.items:
        print("  keys: %s, count: %d" % (str(item.keys), item.row_count))
        for sub_agg in item.sub_aggs:
            print("    sub_agg: %s, value: %s" % (sub_agg.name, str(sub_agg.value)))