teehr.VariableTable#

class teehr.VariableTable(ev)[source]#

Bases: BaseTable

Access methods to variables table.

Methods

`add`	Add a unit to the evaluation.
`distinct_values`	Return distinct values for a column.
`field_enum`	Get the variable fields enum.
`fields`	Return table columns as a list.
`filter`	Apply a filter.
`order_by`	Apply an order_by.
`query`	Run a query against the table with filters and order_by.
`to_geopandas`	Return GeoPandas DataFrame.
`to_pandas`	Return Pandas DataFrame.
`to_sdf`	Return PySpark DataFrame.

add(variable: Variable | List[Variable])[source]#

Add a unit to the evaluation.

Parameters:: variable (Union[Variable, List[Variable]]) – The variable domain to add.

Example

>>> from teehr.models.domain_tables import Variable
>>> variable = Variable(
>>>     name="streamflow_hourly_inst",
>>>     long_name="Instantaneous streamflow"
>>> )
>>> ev.load.add_variable(variable)

distinct_values(column: str) → List[str]#: Return distinct values for a column.

field_enum() → VariableFields[source]#: Get the variable fields enum.

fields() → List[str]#: Return table columns as a list.

Apply a filter.

Parameters:

filters (Union[) – str, dict, FilterBaseModel,: List[Union[str, dict, FilterBaseModel]]

] The filters to apply to the query. The filters can be a string, dictionary, FilterBaseModel or a list of any of these.

Returns:

self (BaseTable or subclass of BaseTable)

Examples

Filters as dictionary:

>>> ts_df = ev.primary_timeseries.filter(
>>>     filters=[
>>>         {
>>>             "column": "value_time",
>>>             "operator": ">",
>>>             "value": "2022-01-01",
>>>         },
>>>         {
>>>             "column": "value_time",
>>>             "operator": "<",
>>>             "value": "2022-01-02",
>>>         },
>>>         {
>>>             "column": "location_id",
>>>             "operator": "=",
>>>             "value": "gage-C",
>>>         },
>>>     ]
>>> ).to_pandas()

Filters as string:

>>> ts_df = ev.primary_timeseries.filter(
>>>     filters=[
>>>         "value_time > '2022-01-01'",
>>>         "value_time < '2022-01-02'",
>>>         "location_id = 'gage-C'"
>>>     ]
>>> ).to_pandas()

Filters as FilterBaseModel:

>>> from teehr.models.filters import TimeseriesFilter
>>> from teehr.models.filters import FilterOperators
>>>
>>> fields = ev.primary_timeseries.field_enum()
>>> ts_df = ev.primary_timeseries.filter(
>>>     filters=[
>>>         TimeseriesFilter(
>>>             column=fields.value_time,
>>>             operator=FilterOperators.gt,
>>>             value="2022-01-01",
>>>         ),
>>>         TimeseriesFilter(
>>>             column=fields.value_time,
>>>             operator=FilterOperators.lt,
>>>             value="2022-01-02",
>>>         ),
>>>         TimeseriesFilter(
>>>             column=fields.location_id,
>>>             operator=FilterOperators.eq,
>>>             value="gage-C",
>>>         ),
>>> ]).to_pandas()

order_by(fields: str | StrEnum | List[str | StrEnum])#

Apply an order_by.

Parameters:: fields (Union[str, StrEnum, List[Union[str, StrEnum]]]) – The fields to order the query by. The fields can be a string, StrEnum or a list of any of these. The fields will be ordered in the order they are provided.
Returns:: self (BaseTable or subclass of BaseTable)

Examples

Order by string:

>>> ts_df = ev.primary_timeseries.order_by("value_time").to_df()

Order by StrEnum:

>>> from teehr.querying.field_enums import TimeseriesFields
>>> ts_df = ev.primary_timeseries.order_by(
>>>     TimeseriesFields.value_time
>>> ).to_pandas()

Run a query against the table with filters and order_by.

In general a user will either use the query methods or the filter and order_by methods. The query method is a convenience method that will apply filters and order_by in a single call.

Parameters:

filters (Union[) – str, dict, FilterBaseModel,
List[Union[str, dict, FilterBaseModel]]

] The filters to apply to the query. The filters can be a string, dictionary, FilterBaseModel or a list of any of these. The filters
order_by (Union[str, List[str], StrEnum, List[StrEnum]]) – The fields to order the query by. The fields can be a string, StrEnum or a list of any of these. The fields will be ordered in the order they are provided.

Returns:

self (BaseTable or subclass of BaseTable)

Examples

Filters as dictionary:

>>> ts_df = ev.primary_timeseries.query(
>>>     filters=[
>>>         {
>>>             "column": "value_time",
>>>             "operator": ">",
>>>             "value": "2022-01-01",
>>>         },
>>>         {
>>>             "column": "value_time",
>>>             "operator": "<",
>>>             "value": "2022-01-02",
>>>         },
>>>         {
>>>             "column": "location_id",
>>>             "operator": "=",
>>>             "value": "gage-C",
>>>         },
>>>     ],
>>>     order_by=["location_id", "value_time"]
>>> ).to_pandas()

Filters as string:

>>> ts_df = ev.primary_timeseries.query(
>>>     filters=[
>>>         "value_time > '2022-01-01'",
>>>         "value_time < '2022-01-02'",
>>>         "location_id = 'gage-C'"
>>>     ],
>>>     order_by=["location_id", "value_time"]
>>> ).to_pandas()

Filters as FilterBaseModel:

>>> from teehr.models.filters import TimeseriesFilter
>>> from teehr.models.filters import FilterOperators
>>>
>>> fields = ev.primary_timeseries.field_enum()
>>> ts_df = ev.primary_timeseries.query(
>>>     filters=[
>>>         TimeseriesFilter(
>>>             column=fields.value_time,
>>>             operator=FilterOperators.gt,
>>>             value="2022-01-01",
>>>         ),
>>>         TimeseriesFilter(
>>>             column=fields.value_time,
>>>             operator=FilterOperators.lt,
>>>             value="2022-01-02",
>>>         ),
>>>         TimeseriesFilter(
>>>             column=fields.location_id,
>>>             operator=FilterOperators.eq,
>>>             value="gage-C",
>>>         ),
>>> ]).to_pandas()

to_geopandas()#: Return GeoPandas DataFrame.

to_pandas()#: Return Pandas DataFrame.

to_sdf()#

Return PySpark DataFrame.

The PySpark DataFrame can be further processed using PySpark. Note, PySpark DataFrames are lazy and will not be executed until an action is called. For example, calling show(), collect() or toPandas(). This can be useful for further processing or analysis, for example,

>>> ts_sdf = ev.primary_timeseries.query(
>>>     filters=[
>>>         "value_time > '2022-01-01'",
>>>         "value_time < '2022-01-02'",
>>>         "location_id = 'gage-C'"
>>>     ]
>>> ).to_sdf()
>>> ts_df = (
>>>     ts_sdf.select("value_time", "location_id", "value")
>>>    .orderBy("value").toPandas()
>>> )
>>> ts_df.head()