tsam is a python package which uses different machine learning algorithms for the aggregation of time series. The data aggregation can be performed in two freely combinable dimensions: By representing the time series by a user-defined number of typical periods or by decreasing the temporal resolution. tsam was originally designed for reducing the computational load for large-scale energy system optimization models by aggregating their input data, but is applicable for all types of time series, e.g., weather data, load data, both simultaneously or other arbitrary groups of time series.
The documentation of the tsam code can be found here.
- flexible handling of multidimensional time-series via the pandas module
- different aggregation methods implemented (averaging, k-means, exact k-medoids, hierarchical, k-maxoids, k-medoids with contiguity), which are based on scikit-learn, or self-programmed with pyomo
- hypertuning of aggregation parameters to find the optimal combination of the number of segments inside a period and the number of typical periods
- novel representation methods, keeping statistical attributes, such as the distribution
- flexible integration of extreme periods as own cluster centers
- weighting for the case of multidimensional time-series to represent their relevance
It is recommended to install tsam within its own environment. If you are no familiar with python environments, plaese consider to read some external documentation. In the following we assume you have a mamba or conda installation. All conda and mamba command are interchangeable.
If you want to prevent any possible dependency conflicts create a new environment using the following command:
mamba create -n tsam_env python pip
Activate an existing or the newly create environment afterward
mamba activate tsam_env
Directly install via pip from pypi as follows:
pip install tsam
or install from conda forge with the following command:
conda install tsam -c conda-forge
Alternatively, clone a local copy of the repository to your computer
git clone https://github.com/FZJ-IEK3-VSA/tsam.git
Change the directory of your shell into the root folder of the repository
cd tsam
For development, it is recommended to install tsam into its own environment using conda e.g.
conda env create --file=requirement.yml
Afterward activate the environment
conda activate tsam_env
Then install tsam via pip as follows
pip install -e .[dev]
In order to use the k-medoids clustering, make sure that you have installed a MILP solver. As default HiGHS is installed and used. Nevertheless, in case you have access to a license we recommend commercial solvers (e.g. Gurobi or CPLEX) since they have a better performance.
In order to setup a virtual environment in Linux, correct the python name in the Makefile and call
make setup_venv
A small example how tsam can be used is decribed as follows
import pandas as pd
import tsam.timeseriesaggregation as tsamRead in the time series data set with pandas
raw = pd.read_csv('testdata.csv', index_col = 0)Initialize an aggregation object and define the length of a single period, the number of typical periods, the number of segments in each period, the aggregation method and the representation method - here duration/distribution representation which contains the minimum and maximum value of the original time series
aggregation = tsam.TimeSeriesAggregation(raw,
noTypicalPeriods = 8,
hoursPerPeriod = 24,
segmentation = True,
noSegments = 8,
representationMethod = "distributionAndMinMaxRepresentation",
distributionPeriodWise = False
clusterMethod = 'hierarchical'
)Run the aggregation to typical periods
typPeriods = aggregation.createTypicalPeriods()Store the results as .csv file
typPeriods.to_csv('typperiods.csv')A first example shows the capabilites of tsam as jupyter notebook.
A second example shows in more detail how to access the relevant aggregation results required for paramtrizing e.g. an optimization.
The example time series are based on a department publication and the test reference years of the DWD.
MIT License
Copyright (C) 2016-2022 Leander Kotzur (FZJ IEK-3), Maximilian Hoffmann (FZJ IEK-3), Peter Markewitz (FZJ IEK-3), Martin Robinius (FZJ IEK-3), Detlef Stolten (FZJ IEK-3)
You should have received a copy of the MIT License along with this program. If not, see https://opensource.org/licenses/MIT
The core developer team sits in the Institute of Energy and Climate Research - Techno-Economic Energy Systems Analysis (IEK-3) belonging to the Forschungszentrum Jülich.
If you want to use tsam in a published work, please kindly cite our latest journal articles:
- Hoffmann et al. (2022):
The Pareto-Optimal Temporal Aggregation of Energy System Models
If you are further interested in the impact of time series aggregation on the cost-optimal results on different energy system use cases, you can find a publication which validates the methods and describes their cababilites via the following link. A second publication introduces a method how to model state variables (e.g. the state of charge of energy storage components) between the aggregated typical periods which can be found here. Finally yet importantly the potential of time series aggregation to simplify mixed integer linear problems is investigated here.
The publications about time series aggregation for energy system optimization models published alongside the development of tsam are listed below:
- Hoffmann et al. (2021):
The Pareto-Optimal Temporal Aggregation of Energy System Models
(open access manuscript to be found here) - Hoffmann et al. (2021):
Typical periods or typical time steps? A multi-model analysis to determine the optimal temporal aggregation for energy system models - Hoffmann et al. (2020):
A Review on Time Series Aggregation Methods for Energy System Models - Kannengießer et al. (2019):
Reducing Computational Load for Mixed Integer Linear Programming: An Example for a District and an Island Energy System - Kotzur et al. (2018):
Time series aggregation for energy system design: Modeling seasonal storage
(open access manuscript to be found here) - Kotzur et al. (2018):
Impact of different time series aggregation methods on optimal energy system design
(open access manuscript to be found here)
This work is supported by the Helmholtz Association under the Joint Initiative "Energy System 2050 A Contribution of the Research Field Energy" and the program "Energy System Design" and within the BMWi/BMWk funded project METIS.
