---
title: AggregatedObservableData (latest version)
description: API reference for qiskit_noise_learning.data.AggregatedObservableData in the latest version of qiskit-noise-learning
source: https://quantum.cloud.ibm.com/docs/en/api/qiskit-noise-learning/generated/data-aggregated-observable-data
---

# qiskit\_noise\_learning.data.AggregatedObservableData

*class* `qiskit_noise_learning.data.AggregatedObservableData(dataset: Dataset)`

[GitHub](https://github.com/Qiskit/qiskit-noise-learning/tree/stable/0.1/qiskit_noise_learning/data/aggregated_observable_data.py)

Bases: [`LeveledData`](/docs/api/qiskit-noise-learning/generated/data-leveled-data "qiskit_noise_learning.data.leveled_data.LeveledData")

Per-path estimates obtained by aggregating [`ObservableData`](/docs/api/qiskit-noise-learning/generated/data-observable-data "qiskit_noise_learning.data.ObservableData").

This class holds a single estimate per observable, obtained by collapsing the per-randomization structure of [`ObservableData`](/docs/api/qiskit-noise-learning/generated/data-observable-data "qiskit_noise_learning.data.ObservableData"). Each observable estimate is labelled by an unbound path and a corresponding fragment depth. A non-negative fragment depth corresponds to labelling by a bound path with that depth, and a fragment depth of `-1` signals that the estimate corresponds to a genuinely unbound path. For a non-negative fragment depth, the estimate value corresponds to the product of all fidelities in the path, and the estimate for a fragment depth of `-1` corresponds to the product of the fidelities in the repeatable fragment.

- Data variables:

  > - `estimate_values`: A 1d float array of per-observable estimates, with dimensions `("observable",)`.
  > - `estimate_std`: A 1d array of standard deviations for the estimates, with dimensions `("observable",)`.
  > - `time_lbs`: A lower bound on the data collection for each observable, with dimensions `("observable",)`.
  > - `time_ubs`: An upper bound on the data collection for each observable, with dimensions `("observable",)`.
  > - `metadata`: A 1d object array of any additional per-observable data, with dimensions `("observable",)`.

- Coordinates:

  > - `unbound_path`: A 1d array of unbound [`Path`](/docs/api/qiskit-noise-learning/generated/sequences-path "qiskit_noise_learning.sequences.Path") instances labelling each observable, with dimensions `("observable",)`.
  > - `fragment_depth`: A 1d array of type `int` specifying the fragment depth associated to the observable. A value of `-1` indicates an estimate of only the `repeatable_fragment` of the path.

**Parameters**

**dataset** – A `Dataset` with the above formatting.

### \_\_init\_\_

`__init__(dataset: Dataset)`

## Methods

|                                                                                                                                                                                            |                                                                   |
| ------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------ | ----------------------------------------------------------------- |
| [`__init__`](#qiskit_noise_learning.data.AggregatedObservableData.__init__ "qiskit_noise_learning.data.AggregatedObservableData.__init__")(dataset)                                        |                                                                   |
| [`filter_time`](#qiskit_noise_learning.data.AggregatedObservableData.filter_time "qiskit_noise_learning.data.AggregatedObservableData.filter_time")(lb, ub)                                | Filter to data gathered within the time bounds.                   |
| [`from_arrays`](#qiskit_noise_learning.data.AggregatedObservableData.from_arrays "qiskit_noise_learning.data.AggregatedObservableData.from_arrays")(unbound\_paths, fragment\_depths, ...) | Instantiate from data specified as arrays in standard containers. |
| [`merge`](#qiskit_noise_learning.data.AggregatedObservableData.merge "qiskit_noise_learning.data.AggregatedObservableData.merge")(other)                                                   | Merge the data from self and other into a single instance.        |

## Attributes

|                                                                                                                                         |                                     |
| --------------------------------------------------------------------------------------------------------------------------------------- | ----------------------------------- |
| [`dataset`](#qiskit_noise_learning.data.AggregatedObservableData.dataset "qiskit_noise_learning.data.AggregatedObservableData.dataset") | The aggregated observable data set. |

### dataset

Type: `Dataset`

The aggregated observable data set.

### from\_arrays

*classmethod* `from_arrays(unbound_paths: list[Path], fragment_depths: list[int], estimate_values: ndarray[float], estimate_std: ndarray[float], time_lbs: ndarray[datetime64], time_ubs: ndarray[datetime64], metadata: ndarray[object] | None = None) → Self`

Instantiate from data specified as arrays in standard containers.

**Parameters**

- **unbound\_paths** – A list of unbound paths (with `fragment_depth=None`).
- **fragment\_depths** – A list of fragment depths, with `-1` indicating the corresponding estimate is in reference to only the repeatable fragment of the corresponding path.
- **estimate\_values** – A 1d array of per-observable estimates.
- **estimate\_std** – A 1d array of standard deviations.
- **time\_lbs** – A 1d array of time lower bounds.
- **time\_ubs** – A 1d array of time upper bounds.
- **metadata** – Any additional data associated with a given observable.

### merge

`merge(other: Self) → Self`

Merge the data from self and other into a single instance.

**Parameters**

**other** – The other data.

**Returns**

A new instance containing both data sets.

### filter\_time

`filter_time(lb: datetime64, ub: datetime64) → Self`

Filter to data gathered within the time bounds.

**Parameters**

- **lb** – The time lower bound (inclusive).
- **ub** – The time upper bound (inclusive).

**Returns**

The time filtered version of self.