Axibase Time Series Database API Client for Python

Axibase Time Series Database Client for Python

The Axibase Time Series Database API Client for Python enables developers to easily read and write statistics and metadata from the Axibase Time Series Database.

API documentation
Client documentation


Install atsd_client with pip:

pip install atsd_client

Or, you can clone the GitHub repository and run:

python install


Connecting to ATSD

To retrieve data from the Axibase Time Series Database (ATSD), establish a connection with the atsd_client module as follows:

    >>> import atsd_client
    >>> from import SeriesService
    >>> conn = atsd_client.connect()

All data needed to connect and authorize ATSD is by default taken from the file.

Initializing the Service

The client provides a set of services for interacting with a particular type of objects in ATSD, for example, Series, Property, Alert, and Message objects as well as with metadata objects such as Entity, Metric, and EntityGroup. An example of creating a service is provided below.

    >>> svc = SeriesService(conn)

Inserting Series Values

To insert series values into ATSD, initialize a Series object and populate it with timestamped values.

    >>> from atsd_client.models import Sample, Series
    >>> series = Series(entity='sensor123', metric='temperature')
    >>> series.add_samples(
            Sample(value=1, time="2016-07-18T17:14:30Z"),
            Sample(value=2, time="2016-07-18T17:16:30Z")
    >>> svc.insert(series)

You can add version information with an optional version argument (here it is assumed that the power metric is versioned).

    >>> from datetime import datetime
    >>> other_series = Series('sensor123', 'power')
    >>> other_series.add_samples(
                Sample(3,, version={"source":"TEST_SOURCE", "status":"TEST_STATUS"})

Querying Series Values

When querying series values from ATSD, you need to specify entity filter, date filter, and series filter.
Forecast, Versioning, Control, and Transformation filters can also be used to filter the resulting Series objects. See the SeriesQuery documentation page for more information.

Series filter: requires specifying the metric name. You can also include type, tags, and exactMatch parameters in this filter to get more specific series objects.

Entity filter: can be set by providing entity name, names of multiple entities, or the name of the entityGroup or entityExpression.

Date filter: can be set by specifying the startDate, endDate, or interval fields. Some combination of these parameters are required. The startDate and endDate fields can be provided either as special words from endTime syntax, an ISO 8601 formatted string, number of milliseconds since 01.01.1970, or a datetime object.

Finally, to get a list of Series objects matching the specified filters, the query method of the service should be used.

    >>> from atsd_client.models import SeriesQuery, SeriesFilter, EntityFilter, DateFilter
    >>> sf = SeriesFilter(metric="temperature")
    >>> ef = EntityFilter(entity="sensor123")
    >>> df = DateFilter(startDate="2016-02-22T13:37:00Z",
    >>> query_data = SeriesQuery(series_filter=sf, entity_filter=ef, date_filter=df)
    >>> result = svc.query(query_data)
    >>> print(result[0]) #picking first Series object

    2016-07-18T17:14:30+00:00             1
    2016-07-18T17:16:30+00:00             2
    metric: temperature
    aggregate: {'type': 'DETAIL'}
    type: HISTORY
    tags: {}
    data: [<Sample v=1, t=1468862070000.0, version=None>, <Sample v=2, t=1468862190000.0, version=None>]
    entity: sensor123

Querying Versioned Series Values

To fetch series values with version information, add the VersionedFilter to the query with the versioned field equal to True. The example demonstrated below also illustrates how milliseconds can be used to set a date filter.

    >>> import time
    >>> from atsd_client.models import VersioningFilter
    >>> cur_unix_milliseconds = int(time.time() * 1000)
    >>> sf = SeriesFilter(metric="power")
    >>> ef = EntityFilter(entity="sensor123")
    >>> df = DateFilter(startDate="2016-02-22T13:37:00Z", endDate=cur_unix_milliseconds)
    >>> vf = VersioningFilter(versioned=True)
    >>> query_data = SeriesQuery(series_filter=sf, entity_filter=ef, date_filter=df, versioning_filter=vf)
    >>> result = svc.query(query_data)
    >>> print(result[0])
               time         value   version_source   version_status
    1468868125000.0           3.0      TEST_SOURCE      TEST_STATUS
    1468868140000.0           4.0      TEST_SOURCE      TEST_STATUS
    1468868189000.0           2.0      TEST_SOURCE      TEST_STATUS
    1468868308000.0           1.0      TEST_SOURCE      TEST_STATUS
    1468868364000.0          15.0      TEST_SOURCE      TEST_STATUS
    1468868462000.0          99.0      TEST_SOURCE      TEST_STATUS
    1468868483000.0          54.0      TEST_SOURCE      TEST_STATUS

Exploring Results

In order to consume the Series object in pandas, a Python data analysis toolkit, you can utilize the built-in to_pandas_series() and from_pandas_series() methods.

    >>> ts = series.to_pandas_series()
    >>> type(ts.index)
    <class 'pandas.tseries.index.DatetimeIndex'>
    >>> print(ts)
    2015-04-10 17:22:24.048000    11
    2015-04-10 17:23:14.893000    31
    2015-04-10 17:24:49.058000     7
    2015-04-10 17:25:15.567000    22
    2015-04-13 14:00:49.285000     9
    2015-04-13 15:00:38            3

Graphing Results

To plot the series with matplotlib, use the built-in plot() method:

    >>> import matplotlib.pyplot as plt
    >>> series.plot()

SQL queries

To perform SQL queries, use query method from SQLService. Returned table will be an instance of DataFrame class.

    >>> sql = SQLService(conn)
    >>> df = sql.query('select * from jvm_memory_free limit 3')
    >>> df
      entity                  datetime        value
    0   atsd  2017-01-20T08:08:45.829Z  949637320.0  45D266DDE38F
    1   atsd  2017-02-02T08:19:14.850Z  875839280.0  45D266DDE38F
    2   atsd  2017-02-02T08:19:29.853Z  777757344.0  B779EDE9F45D

Implemented Methods

Data API

  • Series
    • Insert
    • Query
  • Properties
    • Insert
    • Query
    • Type Query
  • Alerts
    • Query
    • Delete
    • Update
    • History Query
  • Messages
    • Insert
    • Query

Meta API

  • Metrics
    • Get
    • List
    • Update
    • Create or replace
    • Delete
  • Entities
    • Get
    • List
    • Update
    • Create or replace
    • Delete
  • Entity Group
    • Get
    • List
    • Update
    • Create or replace
    • Delete
    • Get entities
    • Add entities
    • Set entities
    • Delete entities