atoti.cube module¶

class atoti.cube.Cube(name, *, java_api, base_table, session)¶

Cube of a Session.

property aggregates_cache: atoti.aggregates_cache.AggregatesCache¶

Aggregates cache of the cube.

Return type: AggregatesCache

create_parameter_hierarchy_from_column(name, column)¶

Create a single-level hierarchy which dynamically takes its members from a column.

Parameters

name (str) – Name given to the created dimension, hierarchy and its single level.
column (Column) – Column from which to take members.

Example

>>> df = pd.DataFrame(
...     {
...         "Seller": ["Seller_1", "Seller_1", "Seller_2", "Seller_2"],
...         "ProductId": ["aBk3", "ceJ4", "aBk3", "ceJ4"],
...         "Price": [2.5, 49.99, 3.0, 54.99],
...     }
... )
>>> table = session.read_pandas(df, table_name="Seller")
>>> cube = session.create_cube(table)
>>> l, m = cube.levels, cube.measures
>>> cube.create_parameter_hierarchy_from_column(
...     "Competitor", table["Seller"]
... )
>>> m["Price"] = tt.value(table["Price"])
>>> m["Competitor price"] = tt.at(
...     m["Price"], coordinates={l["Seller"]: l["Competitor"]}
... )
>>> cube.query(
...     m["Competitor price"],
...     levels=[l["Seller"], l["ProductId"]],
... )
                   Competitor price
Seller   ProductId
Seller_1 aBk3                  2.50
         ceJ4                 49.99
Seller_2 aBk3                  2.50
         ceJ4                 49.99
>>> cube.query(
...     m["Competitor price"],
...     levels=[l["Seller"], l["ProductId"]],
...     condition=l["Competitor"] == "Seller_2",
... )
                   Competitor price
Seller   ProductId
Seller_1 aBk3                  3.00
         ceJ4                 54.99
Seller_2 aBk3                  3.00
         ceJ4                 54.99

create_parameter_hierarchy_from_members(name, members, *, data_type=None, index_measure_name=None)¶

Create a single-level hierarchy with the given members.

It can be used as a parameter hierarchy in advanced analyses.

Parameters

name (str) – The name of hierarchy and its single level.
members (Sequence[Any]) – The members of the hierarchy.
data_type (Optional[DataType]) – The type with which the members will be stored. Automatically inferred by default.
index_measure_name (Optional[str]) – The name of the indexing measure to create for this hierarchy, if any.

create_parameter_simulation(name, *, measures=None, levels=None, base_scenario_name='Base', **kwargs)¶

Create a parameter simulation and its associated measures.

Parameters

name (str) – The name of the simulation. This is also the name of the corresponding table that will be created.
measures (Optional[Mapping[str, Union[float, int, None]]]) – The mapping from the names of the created measures to their default value.
levels (Optional[Sequence[Level]]) – The levels to simulate on.
base_scenario_name (str) – The name of the base scenario.

Example

>>> sales_table = session.read_csv(
...     f"{TUTORIAL_RESOURCES}/sales.csv",
...     table_name="Sales",
...     keys=["Sale ID"],
... )
>>> shops_table = session.read_csv(
...     f"{TUTORIAL_RESOURCES}/shops.csv",
...     table_name="Shops",
...     keys=["Shop ID"],
... )
>>> sales_table.join(shops_table, mapping={"Shop": "Shop ID"})
>>> cube = session.create_cube(sales_table)
>>> l, m = cube.levels, cube.measures

Creating a parameter simulation on one level:

>>> country_simulation = cube.create_parameter_simulation(
...     "Country simulation",
...     measures={"Country parameter": 1.0},
...     levels=[l["Country"]],
... )
>>> country_simulation += ("France crash", "France", 0.8)
>>> country_simulation.head()
                      Country parameter
Country Scenario
France  France crash                0.8

France crash is the name of the scenario.
France is the coordinate at which the value will be changed.
0.8 is the value the Country parameter measure will have in this scenario.

>>> m["Unparametrized turnover"] = tt.agg.sum(
...     sales_table["Unit price"] * sales_table["Quantity"]
... )
>>> m["Turnover"] = tt.agg.sum(
...     m["Unparametrized turnover"] * m["Country parameter"],
...     scope=tt.scope.origin(l["Country"]),
... )
>>> cube.query(m["Turnover"], levels=[l["Country simulation"]])
                      Turnover
Country simulation
Base                961,463.00
France crash        889,854.60

Drilldown to the Country level for more details:

>>> cube.query(
...     m["Unparametrized turnover"],
...     m["Country parameter"],
...     m["Turnover"],
...     levels=[l["Country simulation"], l["Country"]],
... )
                           Unparametrized turnover Country parameter    Turnover
Country simulation Country
Base               France               358,042.00              1.00  358,042.00
                   USA                  603,421.00              1.00  603,421.00
France crash       France               358,042.00               .80  286,433.60
                   USA                  603,421.00              1.00  603,421.00

Creating a parameter simulation on multiple levels:

>>> size_simulation = cube.create_parameter_simulation(
...     "Size simulation",
...     measures={"Size parameter": 1.0},
...     levels=[l["Country"], l["Shop size"]],
... )
>>> size_simulation += (
...     "Going local",
...     None,  # ``None`` serves as a wildcard matching any member value.
...     "big",
...     0.8,
... )
>>> size_simulation += ("Going local", "USA", "small", 1.2)
>>> m["Turnover"] = tt.agg.sum(
...     m["Unparametrized turnover"]
...     * m["Country parameter"]
...     * m["Size parameter"],
...     scope=tt.scope.origin(l["Country"], l["Shop size"]),
... )
>>> cube.query(
...     m["Turnover"],
...     levels=[l["Size simulation"], l["Shop size"]],
... )
                             Turnover
Size simulation Shop size
Base            big        120,202.00
                medium     356,779.00
                small      484,482.00
Going local     big         96,161.60
                medium     356,779.00
                small      547,725.20

Creating a parameter simulation without levels:

>>> crisis_simulation = cube.create_parameter_simulation(
...     "Global Simulation",
...     measures={"Global parameter": 1.0},
... )
>>> crisis_simulation += ("Global Crisis", 0.9)
>>> m["Turnover"] = m["Unparametrized turnover"] * m["Global parameter"]
>>> cube.query(m["Turnover"], levels=[l["Global Simulation"]])
                     Turnover
Global Simulation
Base               961,463.00
Global Crisis      865,316.70

Creating a parameter simulation with multiple measures:

>>> multi_parameter_simulation = cube.create_parameter_simulation(
...     "Price And Quantity",
...     measures={
...         "Price parameter": 1.0,
...         "Quantity parameter": 1.0,
...     },
... )
>>> multi_parameter_simulation += ("Price Up Quantity Down", 1.2, 0.8)
>>> m["Simulated Price"] = (
...     tt.value(sales_table["Unit price"]) * m["Price parameter"]
... )
>>> m["Simulated Quantity"] = (
...     tt.value(sales_table["Quantity"]) * m["Quantity parameter"]
... )
>>> m["Turnover"] = tt.agg.sum_product(
...     m["Simulated Price"],
...     m["Simulated Quantity"],
...     scope=tt.scope.origin(l["Sale ID"]),
... )
>>> cube.query(m["Turnover"], levels=[l["Price And Quantity"]])
                          Turnover
Price And Quantity
Base                    961,463.00
Price Up Quantity Down  923,004.48

Return type: Table

explain_query(*measures, condition=None, include_totals=False, levels=None, scenario='Base', timeout=30)¶

Run the query but return an explanation of how the query was executed instead of its result.

See also

query() for the roles of the parameters.

Return type: QueryAnalysis
Returns: An explanation containing a summary, global timings, and the query plan with all the retrievals.

property hierarchies: atoti._local_hierarchies._LocalHierarchies¶

Hierarchies of the cube.

Return type: ~_LocalHierarchies

property levels: atoti._local_cube._Levels¶

Levels of the cube.

Return type: ~_Levels

property measures: atoti._local_measures._LocalMeasures¶

Measures of the cube.

Return type: ~_LocalMeasures

property name: str¶

Name of the cube.

Return type: str

query(*measures, condition=None, include_totals=False, levels=None, mode='pretty', scenario='Base', timeout=30)¶

Query the cube to retrieve the value of the passed measures on the given levels.

In JupyterLab with the atoti-jupyterlab plugin installed, query results can be converted to interactive widgets with the Convert to Widget Below action available in the command palette or by right clicking on the representation of the returned Dataframe.

Parameters

measures (Union[Measure, QueryMeasure]) – The measures to query.

condition (Union[LevelCondition, MultiCondition, LevelIsInCondition, HierarchyIsInCondition, None]) –

The filtering condition. Only conditions on level equality with a string are supported.

Examples

>>> df = pd.DataFrame(
...     columns=["Continent", "Country", "Currency", "Price"],
...     data=[
...         ("Europe", "France", "EUR", 200.0),
...         ("Europe", "Germany", "EUR", 150.0),
...         ("Europe", "United Kingdom", "GBP", 120.0),
...         ("America", "United states", "USD", 240.0),
...         ("America", "Mexico", "MXN", 270.0),
...     ],
... )
>>> table = session.read_pandas(
...     df,
...     keys=["Continent", "Country", "Currency"],
...     table_name="Prices",
... )
>>> cube = session.create_cube(table)
>>> del cube.hierarchies["Continent"]
>>> del cube.hierarchies["Country"]
>>> cube.hierarchies["Geography"] = [
...     table["Continent"],
...     table["Country"],
... ]
>>> h, l, m = cube.hierarchies, cube.levels, cube.measures

>>> cube.query(
...     m["Price.SUM"],
...     levels=[l["Country"]],
...     condition=l["Continent"] == "Europe",
... )
                         Price.SUM
Continent Country
Europe    France            200.00
          Germany           150.00
          United Kingdom    120.00


>>> cube.query(
...     m["Price.SUM"],
...     levels=[l["Country"], l["Currency"]],
...     condition=(
...         (l["Continent"] == "Europe")
...         & (l["Currency"] == "EUR")
...     ),
... )
                           Price.SUM
Continent Country Currency
Europe    France  EUR         200.00
          Germany EUR         150.00

>>> cube.query(
...     m["Price.SUM"],
...     levels=[l["Country"]],
...     condition=h["Geography"].isin(
...         ("America",), ("Europe", "Germany")
...     ),
... )
                        Price.SUM
Continent Country
America   Mexico           270.00
          United states    240.00
Europe    Germany          150.00

include_totals (bool) –

Whether the returned DataFrame should include the grand total and subtotals. Totals can be useful but they make the DataFrame harder to work with since its index will have some empty values.

Example

>>> cube.query(
...     m["Price.SUM"],
...     levels=[l["Country"], l["Currency"]],
...     include_totals=True,
... )
                                  Price.SUM
Continent Country        Currency
Total                                980.00
America                              510.00
          Mexico                     270.00
                         MXN         270.00
          United states              240.00
                         USD         240.00
Europe                               470.00
          France                     200.00
                         EUR         200.00
          Germany                    150.00
                         EUR         150.00
          United Kingdom             120.00
                         GBP         120.00

levels (Optional[Sequence[~_Level]]) – The levels to split on. If None, the value of the measures at the top of the cube is returned.
scenario (str) – The scenario to query.
timeout (int) – The query timeout in seconds.
mode (Literal[‘pretty’, ‘raw’]) –
The query mode.
- "pretty" is best for queries returning small results:
  - A QueryResult will be returned and its rows will be sorted according to the level comparators.
  Example:
  >>> cube.query( ... m["Price.SUM"], ... levels=[l["Continent"]], ... mode="pretty", ... ) Price.SUM Continent America 510.00 Europe 470.00
- "raw" is best for benchmarks or large exports:
- A faster and more efficient endpoint reducing the data transfer from Java to Python will be used.
- A classic pandas.DataFrame will be returned.
- include_totals="True" will not be allowed.
- The Convert to Widget Below action provided by the atoti-jupyterlab plugin will not be available.
Example:
>>> cube.query( ... m["Price.SUM"], ... levels=[l["Continent"]], ... mode="raw", ... ) Continent Price.SUM 0 Europe 470.0 1 America 510.0

Return type

Union[QueryResult, DataFrame]

property schema: Any¶

Schema of the cube’s tables as an SVG graph.

Note

Graphviz is required to display the graph. It can be installed with Conda: conda install graphviz or by following the download instructions.

Return type: Any
Returns: An SVG image in IPython and a Path to the SVG file otherwise.

property shared_context: atoti.cube.CubeContext¶

Context values shared by all the users.

Context values can also be set at query time, and per user, directly from the UI. The values in the shared context are the default ones for all the users.

queriesTimeLimit

The number of seconds after which a running query is cancelled and its resources reclaimed. Set to -1 to remove the limit. Defaults to 30s.
queriesResultLimit.intermediateSize

The limit number of point locations for a single intermediate result. This works as a safe-guard to prevent queries from consuming too much memory, which is especially useful when going to production with several simulatenous users on the same server. Set to -1 to use the maximum limit. In atoti, the maximum limit is the default while in Atoti+ it defaults to 1000000.
queriesResultLimit.tansientResultSize

Similar to intermediateSize but across all the intermediate results of the same query. Set to -1 to use the maximum limit. In atoti, the maximum limit is the default while in Atoti+ it defaults to 10000000.

Example

>>> df = pd.DataFrame(
...     columns=["City", "Price"],
...     data=[
...         ("London", 240.0),
...         ("New York", 270.0),
...         ("Paris", 200.0),
...     ],
... )
>>> table = session.read_pandas(
...     df, keys=["City"], table_name="shared_context example"
... )
>>> cube = session.create_cube(table)
>>> cube.shared_context["queriesTimeLimit"] = 60
>>> cube.shared_context["queriesResultLimit.intermediateSize"] = 1000000
>>> cube.shared_context["queriesResultLimit.transientSize"] = 10000000
>>> cube.shared_context
{'queriesTimeLimit': '60', 'queriesResultLimit.intermediateSize': '1000000', 'queriesResultLimit.transientSize': '10000000'}

Return type: CubeContext

class atoti.cube.CubeContext(_java_api, _cube)¶