Skip to main content

(time series model analyses) This package offers tools for analyzing models that generate multivariate time series as output.

Project description

tsma (time series model analyses)

This package offers tools for analyzing models that generate multivariate time series as output. The goal is to make the study of large models, especially those related to macroeconomic ABMs, more convenient.

References:

This package contains an implementation of two papers:

To learn more about signature transform see Adeline Fermanian's thesis of (2021) or https://arxiv.org/pdf/1911.13211.pdf

Structure:

model

  • model : model class that serves as baseline
  • gross2022 : implementation of Gross's model of 2022

collect

  • output_management : This package includes functions for encoding and decoding parameters into databases, as well as storing and managing simulation data.
  • iterators : functions used to gather simulations results
  • data_collect : allows for the exploration of parameters through statistical analyses, with the option for automatic saving

visuals

  • fig_constructors : assists in the creation of figures using Plotly, Seaborn, and Networkx
  • figures : creates figures using Plotly, Seaborn, and Networkx
  • fig_management : manages the creation and saving of multiple figures
  • dashboards : for creating a dashboard on a local server, which allows for the visualization of a summary of a simulation with various sets of parameters.

analyses

  • statistics : Means, Stationarity tests, ...
  • vandin : implementation of statistical analyses taken from : Vandin & al (2021)
  • metrics : classes used to create clustering approaches for time series clustering
  • clustering :
    • multivariate time series unsupervised clustering
    • clustering scoring and selection
    • phase diagram visualization ( parameters space projection of clustering results )
    • comparison of clusters

basics

  • transfers : functions used to divide and join strings, lists and dictionaries
  • text_management : name conversions, title creations ...

Tutorials:

Find tutorials at : https://github.com/S-bazaz/tsma

  • tsma1 : how to format your model and use the saving system (save, overwrite, query ...)
  • tsma2 : how to use visualizations ( simple plots, dashboards ...)
  • tsma3 : vandin & al statistical checking and data collection (transient analysis, parameters exploration ...)
  • tsma4 : time series clustering ( clusters computation, clustering selection, phase diagram ...)
  • tsma5 : clustering comparison ( visualization per cluster, dendrograms, Jaccard coefficients network plots ...)

Project details


Download files

Download the file for your platform. If you're not sure which to choose, learn more about installing packages.

Source Distribution

tsma-1.2.0.tar.gz (69.3 kB view hashes)

Uploaded Source

Built Distribution

tsma-1.2.0-py3-none-any.whl (75.7 kB view hashes)

Uploaded Python 3

Supported by

AWS AWS Cloud computing and Security Sponsor Datadog Datadog Monitoring Fastly Fastly CDN Google Google Download Analytics Microsoft Microsoft PSF Sponsor Pingdom Pingdom Monitoring Sentry Sentry Error logging StatusPage StatusPage Status page