> For the complete documentation index, see [llms.txt](https://time-series-features.gitbook.io/pyspi/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/pyspi/information-about-pyspi/api-reference/pyspi.data.data.md).

# pyspi.data.Data

> *<mark style="color:blue;">class</mark>* **pyspi.data.Data**<mark style="color:blue;">(</mark>*<mark style="color:blue;">data=None</mark>*<mark style="color:blue;">,</mark> *<mark style="color:blue;">dim\_order='ps'</mark>*<mark style="color:blue;">,</mark> *<mark style="color:blue;">detrend=False</mark>*<mark style="color:blue;">,</mark> *<mark style="color:blue;">normalise=True</mark>*<mark style="color:blue;">,</mark> *<mark style="color:blue;">name=None</mark>*<mark style="color:blue;">,</mark> *<mark style="color:blue;">procnames=None</mark>*<mark style="color:blue;">,</mark> *<mark style="color:blue;">n\_processes=None</mark>*<mark style="color:blue;">,</mark> *<mark style="color:blue;">n\_observations=None</mark>*<mark style="color:blue;">)</mark>

**Store data for dependency analysis.**

Data takes a 2-dimensional array representing measurements in two dimensions: processes (p) and observations (s). The order of the dimensions in the provided array are specified using a two-character string, either '*ps*' for an array with realisations over (dim 1) processes, and (dim 2) observations in time, or '*sp*', denoting an array over (dim 1) observations in time, and (dim 2) processes.

### Example

```python
# Initialise empty data object
data = Data()

# Load a prefilled financial dataset
data_forex = Data().load_dataset(forex)

# Create data objects with data of various sizes
d = np.arange(3000).reshape((3, 1000))  # 3 procs.,
data_2 = Data(d, dim_order='ps')        # 1000 observations

# Overwrite data in existing object with random data
d = np.arange(5000)
data_2.set_data(data_new, 's')
```

***

<table><thead><tr><th width="190">Parameters:</th><th>Description</th></tr></thead><tbody><tr><td></td><td><ul><li><strong>data</strong> (<em>array_like, optional</em>) – 2-dimensional array with raw data, default=None.</li><li><strong>dim_order</strong> (<a href="https://docs.python.org/3/library/stdtypes.html#str"><em>str</em></a><em>, optional</em>) – Order of dimensions, accepts two combinations of the characters ‘p’, and ‘s’ for processes and observations, default=ps’ (process along the first axis and observation along the second axis).</li><li><strong>detrend</strong> (<a href="https://docs.python.org/3/library/functions.html#bool"><em>bool</em></a><em>, optional</em>) - If True, detrend the dataset along the time axis before normalising (if enabled), default=True.</li><li><strong>normalise</strong> (<a href="https://docs.python.org/3/library/functions.html#bool"><em>bool</em></a><em>, optional</em>) – If True, z-score normalise the dataset along the time axis before computing SPIs, default=True.</li><li><strong>name</strong> (<a href="https://docs.python.org/3/library/stdtypes.html#str"><em>str</em></a><em>, optional</em>) – Name of the dataset</li><li><strong>procnames</strong> (<a href="https://docs.python.org/3/library/stdtypes.html#list"><em>list</em></a><em>, optional</em>) – List of process names with length the number of processes, default=None.</li><li><strong>n_processes</strong> (<a href="https://docs.python.org/3/library/functions.html#int"><em>int</em></a><em>, optional</em>) – Truncates data to this many processes, default=None.</li><li><strong>n_observations</strong> (<a href="https://docs.python.org/3/library/functions.html#int"><em>int</em></a><em>, optional</em>) – Truncates data to this many observations, default=None.</li></ul></td></tr></tbody></table>

> `__init__`(*data=None*, *dim\_order='ps'*, *detrend=False,* *normalise=True*, *name=None*, *procnames=None*, *n\_processes=None*, *n\_observations=None*)

## Methods

| Method                                                                                                                                                          | Description                                                      |
| --------------------------------------------------------------------------------------------------------------------------------------------------------------- | ---------------------------------------------------------------- |
| [`__init__`](https://pyspi-toolkit.readthedocs.io/en/latest/generated/pyspi.data.Data.html#pyspi.data.Data.__init__)(\[data, dim\_order, normalise, name, ...]) | -                                                                |
| `add_process`(proc\[, verbose])                                                                                                                                 | Appends a univariate process to an existing dataset.             |
| `convert_to_numpy`(data)                                                                                                                                        | Converts other data instances to default numpy format.           |
| `remove_process`(procs)                                                                                                                                         | Remove a univariate process from an existing dataset.            |
| `set_data`(data\[, dim\_order, name, ...])                                                                                                                      | Overwrite data in an existing Calculator instance with new data. |
| `to_numpy`(\[realisation, squeeze])                                                                                                                             | Return the dataset as a numpy array.                             |

## Attributes

| Attribute   | Description              |
| ----------- | ------------------------ |
| `name`      | Name of the data object. |
| `procnames` | List of process names.   |
|             |                          |


---

# 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/pyspi/information-about-pyspi/api-reference/pyspi.data.data.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.