Changelog

0.13.8

Dependencies

  • Upgrade python-dateutil latest from <=2.9 to <2.10

0.13.7

Dependencies

  • Upgrade dateutil latest from 2.8 to 2.9

0.13.6

Improvements

  • Add issued_time_of_day parameter to eq.instances.list(), eq.instances.load() and eq.instances.latest() for filtering instances based on issued time

0.13.5

Bugfixes

  • Adjust date formatting in AbsoluteResult.to_df() to be consistent with the rest of the client

0.13.4

Improvements

  • Add AbsoluteResult.to_df() and AbsoluteResult.to_dataframe()

Bugfixes

  • The kind parameter on eq.metadata.places() did nothing. It is now fixed.

0.13.3

Improvements

  • Modify the User-Agent header to comply with standard conventions

0.13.2

Bugfixes

  • Fix deadlock in push feed preventing the client to resubscribe after recovery from a connection loss

0.13.1

Improvements

  • Add new border:

    • DK1 – GB Explicit

Bugfixes

  • Remove call to logging.basicConfig() in energyquantified/api/events.py as it was causing issues with the logging configuration in the client’s parent application

0.13

Improvements

  • Parse Instance’s created and modified in the Curve’s time zone.

  • Add eq.instances.rolling() for rolling forecasts

0.12.1

Bugfixes

  • Remove code setting default log level to DEBUG

  • Fix a bug introduced in v0.12 that caused parsing of curve events to fail

0.12

Improvements

  • Add PeriodInstancesAPI.relative() for relative queries

  • Add modified-at-latest parameter for eq.instances.relative()

  • Implement absolute forecasts for instances eq.instances.absolute()

  • Add unit and denominator attributes to Series and OHLCList

  • Add support for unit conversion when loading data from the API

  • Add curve_type parameter for eq.metadata.curves()

  • Add Subscription, SubscriptionAccess, SubscriptionType and SubscriptionCollectionPerm models

  • Add Curve.subscription field, providing the user with subscription information for curves

  • Add User, Organization and AccountManager models

  • Add UserAPI and eq.user.user() to get details of the current user

Bugfixes

  • Remove unnecessary limitation from eq.instances.relative() for parameters time_of_day, after_time_of_day and before_time_of_day

Breaking changes

  • Removed Place.area. Use Place.areas instead.

0.11

Improvements

  • Implement EventsAPI for streaming events from Energy Quantified’s WebSocket API (push feed).

  • Add timezone conversion.

  • Add threshold parameter to define how many values are allowed to be missing while performing an aggregation.

Dependencies

  • Add websocket-client v1.5.1

  • Upgrade requests to minimum 2.31 due to security fixes

0.10.1

Bugfixes

  • Fixed a bug introduced in v0.10 in PeriodSeries.to_timeseries()

0.10

Improvements

  • Add areas IS and LU

  • Update borders:

    • Set SI – HU and SI – HU as Flow-based

  • Increase rate limits.

  • Increase default request timeout.

  • Add option to include proxies in EnergyQuantified and RealtoConnection.

  • Add support for using periods installed capacity instead of values when converting a PeriodSeries to a Timeseries or a DataFrame.

Bugfixes

  • Fixed an error where SRMC responses failed to parse empty lists in the response

0.9.1

Improvements

  • Improve Border.__hash__, Border.__eq__ and Border.__ne__ methods

  • Add new areas:

    • MA Morocco

    • LY Libya

    • DZ Algeria

  • Add new border:

    • SI – HU Flow-based

  • Borders that no longer has commercial capacity:

    • LV – RU

    • FI – RU

    • LT – RU_KGD

  • Borders that have changed to flow-based:

    • AT – SI

    • HR – HU

    • CZ – DE

    • CZ – PL

    • HU – SK

    • PL – SK

    • PL – DE

    • CZ – SK

    • HU – RO

  • Border updated to flowed-based and explicit:

    • SI – HR

0.9

Bugfixes

  • When invoking timeseries.to_dataframe(single_level_header=True) the resulting column index was still a MultiIndex but with a single level. Now the resulting column index is a normal Index type.

0.8.1

Dependencies

  • requests: Use the latest v2.x available, as requests is very stable library.

  • python-dateutil: Use the latest v2.8.x available.

0.8

Improvements

  • Add eq.metadata.curve() which returns a Curve object for the corresponding curve name.

  • Add a section in the metadata documentation on the eq.metadata.curve() method.

  • Add eq.metadata.curves() and eq.metadata.curve() to reference page in the documentation.

  • Add Area.short_tag which is a shorter tag than Area.tag. It is typically used for TSO areas. Example: The area with tag DE-Amprion has the short tag Amprion.

  • Changes in capacity allocation for these borders:

    • AT-CZ Implicit only

    • AT-HU Implicit only

    • NO2-GB Implicit

  • Add borders:

    • RS-XK Explicit

    • TR-GE Explicit

0.7.1

More gas data preparations.

Improvements

  • Add Place.areas (list of areas), as some places (i.e. gas interconnectors) are places on borders and should be listed for both areas.

  • Add three new place types: PlaceType.GAS_STORAGE, PlaceType.GAS_LNG_TERMINAL and PlaceType.GAS_INTERCONNECTOR

  • Add Curve.commodity (str) which is either Power, Gas, Coal, Oil, Carbon or None at this time.

  • Add a commodity filter for eq.metadata.curves().

Deprecations

  • Add Place.area is deprecated and will eventually be replaced by Place.areas. It will be removed in a future release.

0.7

Implementing Acer’s non-standard Gas Day timezone.

Improvements

  • Define new timezone in pytz called Europe/Gas_Day. It follows Acer’s Gas Day, which is from 06:00 – 06:00 in CET/CEST. This timezone is used for the natural gas market in the European Union. Import it with from energyquantified.time import GAS_DAY, or look it up in pytz like so: pytz.timezone("Europe/Gas_Day").

Bugfixes

  • When invoking timeseries.to_dataframe(name="foo", single_level_header=True), the resulting column header in pandas’ DataFrame no longer includes the instance identifier. However, the ensemble/scenario name is still appended at the end.

0.6.3

Improvements

  • Increase rate limits.

Bugfixes

  • Set has_instances = True in CurveType.INSTANCE_PERIOD (was False).

0.6.2

Improvements

  • Add new area Kosovo (Area.XK) with these borders:

    • XK–AL Explicit

    • XK–ME Explicit

    • XK–MK Explicit

  • Add new border:

    • NO2–GB Explicit

  • Changes in capacity allocation for these borders:

    • IT-Sud–GR Implict and Explicit

    • BG–GR Implict and Explicit

    • PL–DE Implict and Explicit

    • PL–SK Implict and Explicit

    • PL–CZ Implict and Explicit

    • DE–CZ Implict and Explicit

    • AT–CZ Implict

    • AT–HU Implict

    • GB–FR Explicit

    • GB–BE Explicit

    • GB–NL Explicit

  • Remove border:

    • RS–AL

0.6.1

Improvements

  • Add ContractPeriod.WEEKEND for OHLC data.

Bugfixes

  • Fix crashes in Border.__str__ and Border.__repr__ due to missing implementations of __lt__ and __gt__ in class Allocation (thanks to stanton119).

0.6

A release with lots of small improvements.

Improvements

  • Add borders and parent-child relationships for the Italian price zone Calabria. The price zone has been in the client for a while, but haven’t placed in the exchange neighbour list for the other price zones in Italy until now.

  • Add the new parameter single_level_header to all to_dataframe() methods. By default, the to_dataframe()-method will create pandas.DataFrame objects with three column headers. When single_level_header=True, the client will merge all three levels into one header. The parameter defaults to False (to not break the old behaviour).

  • Remove the parameter hhv_to_lhv for all SRMC API operations.

  • Add a new class RealtoConnection. This class is a drop-in replacement for the EnergyQuantified-class. It lets Realto users connect to the Energy Quantified’s API on Realto’s marketplace.

  • Update the documentation on how to authenticate for Realto users.

  • Add a quickstart chapter for Realto users.

  • Add a section in the pandas documentation on the effects of setting the single_level_header parameter to True in to_dataframe().

  • Add documentation on the fill parameter in eq.ohlc.load_delivery_as_timeseries() and eq.ohlc.load_front_as_timeseries().

  • Other minor improvements in the documentation.

Breaking change

  • Remove the HHV-to-LHV option for gas in the SRMC API.

Bugfixes

  • Slashes (/) weren’t escaped in curve names in the URL. While this didn’t cause issues for Energy Quantified’s API, it caused an issue while integrating the client with Realto’s marketplace.

Dependencies

  • Upgrade requests to v2.25.1.

0.5

Introducing support for short-run marginal cost (SRMC) calculations from OHLC data.

Improvements

  • Add OhlcAPI#latest_as_periods() method for generating a “forward curve” from all closing prices in a market.

  • Add fill parameter to OhlcAPI#load_front_as_timeseries() and OhlcAPI#load_front_as_timeseries().

  • Add SRMC and SRMCOptions data classes.

  • Implement the SRMC API: load_front(), load_delivery(), load_front_as_timeseries(), load_delivery_as_timeseries(), latest(), and latest_as_periods().

  • Add section in the OHLC documentation on how to load “forward curves”.

  • Add new chapter on SRMC in the documentation.

Bugfixes

  • Fix a crash in the Contract JSON parser that occured only for SRMC operations.

Dependencies

  • Upgrade requests to v2.25.0.

0.4.2

Improvements

  • Update border configurations (such as the AELGrO cable between Belgium and Germany, for instance).

Bugfixes

  • Add missing area (SEM).

0.4.1

Bugfixes

  • Fix a crash in TimeseriesList#to_dataframe().

0.4

Improve pandas integration with more utility methods.

Improvements

  • Page objects are now immutable (for curve and place search responses).

  • Add Series.set_name() to let users set a custom name for time series’ and period-based series’.

  • Add TimeseriesList with a to_dataframe() method for converting a list of time series to a pandas data frame. It subclasses Python’s built-in list and overrides its methods with extra validations.

  • Add PeriodseriesList. Similar to TimeseriesList, it subclasses Python’s list. It has two methods: (1) to_timeseries() which converts this list to a TimeseriesList, and (2) to_dataframe(frequency) which converts this list to a data frame.

  • Add Periodseries#to_dataframe(frequency). Previously, you would have to first convert the period-based series to a time series and then call to_dataframe.

  • Update headers in pandas data frames.

  • Add OHLCList#to_dataframe() for converting OHLC data to a data frame.

  • Update documentation where applicable with a short description on how to convert time series, period-based series and OHLC data to data frames.

  • Add own chapter on how to convert data to pandas.DataFrame.

  • Add own chapter on packages and where to find the different classes and enumerators.

Breaking change

With better pandas integration, we changed column headers for data frames. As of v0.4, data frames have three column header levels for time series data:

  1. Curve name

  2. Instance or contract

  3. Scenario (ensemble)

We did this to better describe the data when converted from time series’ to pandas data frames. Refer to the chapter on pandas integration for more details.

0.3

Introducing support for OHLC data (open, high, low, close).

Improvements

  • Implement operations in the OHLC API: load(), latest(). load_delivery_as_timeseries(), and load_front_as_timeseries()

  • Add data and metadata classes for OHLC: OHLCField, ContractPeriod, Product, OHLC, OHLCList, and Contract.

  • Add member Series#contract, which is a reference to a set by the load_*_as_timeseries()-operations.

  • Add documentation for OHLC.

  • Add new curve data type: DataType.SCENARIO.

Bugfixes

  • Fix runtime error in Series#name() (Series is superclass of Timeseries and Periodseries).

  • ValidationError exceptions occuring on the server-side didn’t include which parameter that failed due to a bug in the JSON error message parser.

0.2

A small release with two improvements.

Improvements

  • Add Periodseries#print() method.

  • Increase 1-10 days-ahead constraints for relative queries to 0-10000.

0.1

The first public release of Energy Quantified’s Python client. Woho!

Improvements

  • Add utilities for working with date-times, frequencies, timezones and resolutions.

  • Add metadata classes for areas, curves, instances, places and more.

  • Add classes for time series and period-based series.

  • Add wrapper around requests with rate-limiting, auto-retry on failure and authentication.

  • Implement APIs for metadata, timeseries, instances, periods and period-instances.

  • Add support for timeseries-to-pandas conversion.

  • Add meaningful exceptions.

  • Add a few examples to the git repo.

  • Write tons of documentation.

Dependencies

  • Add pytz, tzlocal, python-dateutil, requests.

  • Not adding pandas, as it is optional.

Bugfixes

  • (None in this release, but probably introduced some!)