> For the complete documentation index, see [llms.txt](https://time-series-features.gitbook.io/hctsa-manual/llms.txt). Markdown versions of documentation pages are available by appending `.md` to page URLs; this page is available as [Markdown](https://time-series-features.gitbook.io/hctsa-manual/information-about-hctsa/faq.md).

# FAQ

Frequently asked questions pertaining to the use of hctsa.

Select a drop-down for more information:

<details>

<summary><strong>How can I use more machine-learning algorithms rather than SVM linear only using <code>TS_Classify</code>?</strong></summary>

You can specify the classification algorithm by setting `cfnParams.whatClassifier` and passing `cfnParams` to `TS_Classify` (the default, set by `GiveMeDefaultClassificationParams`, is `'svm-linear-lowdim'`). You can see the options in `GiveMeCfn`. Typically because there is complexity in the embedding in a high-dimensional feature space, we have tried to remove complexity in the classifiers (to avoid overfitting), also for interpretability. You can also export the features using `OutputToCSV` and analyze them in other environments (like Python).

</details>

<details>

<summary><strong>Can I perform a multivariate time series analysis using </strong><em><strong>hctsa</strong></em><strong>?</strong></summary>

*hctsa* is designed to extract thousands of features from a ***single univariate time series***; [pyspi](https://github.com/DynamicsAndNeuralSystems/pyspi) implements hundreds of pairwise dependence measures from multivariate time series.

You could incorporate both (i) univariate features of the components of the system; and (ii) features summarizing the pairwise dependence structure of the distributed system.

For example, some points to consider:

* You could compute univariate features of each component of your system and then concatenate them to combine features of all individual time series:
  * To do this you may consider using a reduced set of features, like [catch22](https://github.com/DynamicsAndNeuralSystems/catch22) to avoid massive dimensionality explosion.
  * You may alternatively use *hctsa* and then do feature selection or dimensionality reduction to tailor a reduced feature set to your application.
  * You may also consider adding some pairwise dependence measures to summarize the multivariate structure, cf. [pyspi](https://github.com/DynamicsAndNeuralSystems/pyspi).
* You could compute univariate features of extracted dominant components of your multivariate system, e.g., using PCA.

</details>

<details>

<summary><strong>How can I export the extracted features?</strong></summary>

Use `OutputToCSV`. This gives you `.csv` files corresponding to a given *hctsa* calculation that you can analyze however you please.

</details>

<details>

<summary><strong>Can I use </strong><em><strong>hctsa</strong></em><strong> features without MATLAB (e.g., in Python, R or Julia)?</strong></summary>

There are a few options for working without MATLAB:

* Use the Python package [*pyhctsa*](https://github.com/DynamicsAndNeuralSystems/pyhctsa), where future development is planned to happen.
* Use an alternative (e.g., native R or native Python) feature-extraction tool, such as those listed [here](/hctsa-manual/information-about-hctsa/related-time-series-resources.md). These have far fewer features than *hctsa* but can get you some of the way there.
* Use a reduced feature set coded in C, which can be used efficiently from many programming languages: [*catch22*](https://github.com/DynamicsAndNeuralSystems/catch22) (22 features selected for general time-series classification) or [*catchaMouse16*](https://github.com/DynamicsAndNeuralSystems/catchaMouse16) (16 features selected for fMRI). You can also derive a reduced set of useful features for your own problem with *hctsa*, and then implement them (or find non-MATLAB implementations).

Note also the licenses: the *hctsa* framework for running analyses and visualizations is licensed as [CC BY-NC-SA 4.0](https://creativecommons.org/licenses/by-nc-sa/4.0/) (a commercial license is available from [Engine Analytics](http://www.engineanalytics.org/)), and the code for computing features is licensed as [GPL v3](https://www.gnu.org/licenses/gpl-3.0.en.html).

</details>

***


---

# Agent Instructions
This documentation is published with GitBook. GitBook is the documentation platform designed so that both humans and AI agents can read, navigate, and reason over technical content effectively. Learn more at gitbook.com.

## Querying This Documentation
If you need additional information that is not directly available in this page, you can query the documentation by asking a question.

Perform an HTTP GET request on the following URL with the `ask` and `goal` query parameters:

```
GET https://time-series-features.gitbook.io/hctsa-manual/information-about-hctsa/faq.md?ask=<question>&goal=<user_goal>
```

`ask` is the immediate question: it should be specific, self-contained, and written in natural language.
`goal` is what the user is ultimately trying to achieve, the reason they need the answer. Sharing it helps GitBook give you a better, more relevant answer. A goal is most helpful when it describes the outcome the user wants rather than restating the question. For example, with `ask=how do I create an API token`, a goal like `automate deployments from our CI pipeline` lets GitBook tailor the answer to that use case.

The response will contain a direct answer to the question and relevant excerpts and sources from the documentation.

Use this mechanism when the answer is not explicitly present in the current page, you need clarification or additional context, or you want to retrieve related documentation sections.