About ebcpy

ebcpy

This PYthon package provides generic functions and classes commonly used for the analysis and optimization of energy systems, buildings and indoor climate (EBC).

Key features are:

TimeSeriesData
SimulationAPI‘s
Optimization wrapper
Pre-/Postprocessing
Modelica utilities

It was developed together with AixCaliBuHA, a framework for an automated calibration of dynamic building and HVAC models. During this development, we found several interfaces relevant to further research. We thus decoupled these interfaces into ebcpy and used the framework, for instance in the design optimization of heat pump systems (link).

Installation

To install, simply run

pip install ebcpy

In order to use all optional dependencies (e.g. pymoo optimization), install via:

pip install ebcpy[full]

If you encounter an error with the installation of scikit-learn, first install scikit-learn separatly and then install ebcpy:

pip install scikit-learn
pip install ebcpy

If this still does not work, we refer to the troubleshooting section of scikit-learn: https://scikit-learn.org/stable/install.html#troubleshooting. Also check issue 23 for updates.

In order to help development, install it as an egg:

git clone https://github.com/RWTH-EBC/ebcpy
pip install -e ebcpy

How to get started?

We recommend running our jupyter-notebook to be guided through a helpful tutorial.
For this, run the following code:

# If jupyter is not already installed:
pip install jupyter
# Go into your ebcpy-folder (cd \path_to_\ebcpy) or change the path to tutorial.ipynb and run:
jupyter notebook tutorial\tutorial.ipynb

Or, clone this repo and look at the examplesREADME.md file. Here you will find several examples to execute.

How to cite ebcpy

Please use the following metadata to cite ebcpy in your research:

@article{Wuellhorst2022,
  doi = {10.21105/joss.03861},
  url = {https://doi.org/10.21105/joss.03861},
  year = {2022},
  publisher = {The Open Journal},
  volume = {7},
  number = {72},
  pages = {3861},
  author = {Fabian Wüllhorst and Thomas Storek and Philipp Mehrfeld and Dirk Müller},
  title = {AixCaliBuHA: Automated calibration of building and HVAC systems},
  journal = {Journal of Open Source Software}
}

TimeSeriesData

Note that we use our own TimeSeriesData object which inherits from pd.DataFrame. The aim is to make tasks like loading different filetypes or applying multiple tags to one variable more convenient, while conserving the powerful tools of the DataFrame. Just a quick intro here:

Variables and tags

>>> from ebcpy.data_types import TimeSeriesData
>>> tsd = TimeSeriesData(r"path_to_a_supported_file")
>>> print(tsd)
Variables    T_heater              T_heater_1
Tags             meas         sim        meas         sim
Time
0.0        313.165863  313.165863  293.173126  293.173126
1.0        312.090271  310.787750  293.233002  293.352448
2.0        312.090027  310.796753  293.385925  293.719055
3.0        312.109436  310.870331  293.589233  294.141754

As you can see, our first column level is always a variable, and the second one a tag. This is especially handy when dealing with calibration or processing tasks, where you will have multiple versions (tags) for one variable. The default tag is raw to indicate the unmodified data. To access a variable, you have to call .loc. To access multiple variables that all hold one tag use xs:

# All tags:
tsd.loc[:, "variable_name"]
# One specific tag:
tsd.loc[:, ("variable_name", "tag_name")]
# One tag, all variables:
tsd.xs("tag_name", axis=1, level=1)

FloatIndex and DateTimeIndex

Measured data typically holds a datetime stamps (DateTimeIndex) while simulation result files hold absolute seconds (FloatIndex). You can easily convert back and forth using:

# From Datetime to float
tsd.to_float_index()
# From float to datetime
tsd.to_datetime_index()
# To clean your data and create a common frequency:
tsd.clean_and_space_equally(desired_freq="1s")