Echopype

https://echopype.readthedocs.io/en/latest/
Admin
Echopype is a package built to enable interoperability and scalability in ocean sonar data...
Tokens:161,506
Snippets:1,244
Trust Score:9.6
Update:4 months ago
Show doc for...
Tokens:
Raw
### Channel Index Example

Source: https://echopype.readthedocs.io/en/latest/data-format-raw.html

Example of PandasIndex for channels.

```python
PandasIndex(Index(['WBT 714581-15 ES18', 'WBT 714583-15 ES120-7C', 'WBT 714590-15 ES70-7C',
       'WBT 714596-15 ES38-7', 'WBT 714597-15 ES333-7C',
       'WBT 714605-15 ES200-7C'],
      dtype='object', name='channel'))
```

--------------------------------

### Install and Run Pre-commit Hooks

Source: https://echopype.readthedocs.io/en/latest/_sources/contrib_setup.md

Install pre-commit hooks for local development to automatically check code formatting and style before committing. Run all hooks across all files to ensure compliance.

```shell
pre-commit install
```

```shell
pre-commit run --all-files
```

--------------------------------

### Example Sensor Configuration

Source: https://echopype.readthedocs.io/en/latest/data-format-raw.html

This snippet shows an example of a sensor configuration within an XML structure, specifying sensor details and telegram values.

```xml
<Configuration>
  <ConfiguredSensors>
    <Sensor Name="ADCP" Type="ADCP">
      <Telegram Name="WaterVelocity" Priority="1" />
      <Telegram Name="BottomVelocity" Priority="1" />
      <Telegram Name="WaterTemperature" Priority="1" />
      <Telegram Name="Heave" Priority="1" />
    </Sensor>
  </ConfiguredSensors>
</Configuration>
```

--------------------------------

### Install Pre-commit Hooks

Source: https://echopype.readthedocs.io/en/latest/contrib_setup.html

Install pre-commit hooks to automatically check and format code before committing. This ensures code style consistency.

```bash
pre-commit install
```

--------------------------------

### Simrad Transmit Frequency Start Example

Source: https://echopype.readthedocs.io/en/latest/data-format-raw.html

Example of the start frequency in transmitted pulses for Simrad echosounders. Units are in Hertz (Hz).

```python
array([1.8e+04, 1.8e+04, 2e+05, 2e+05])
```

--------------------------------

### EK60 Time Data Example

Source: https://echopype.readthedocs.io/en/latest/data-format-raw.html

Example of time data extracted from EK60 sonar, showing a list of datetime objects.

```python
array(['2017-07-28 18:16:30.586572800',
               '2017-07-28 18:16:33.394732032',
               '2017-07-28 18:16:36.211893248',
               '2017-07-28 18:16:39.030055936',
               '2017-07-28 18:16:41.850217984',
               '2017-07-28 18:16:44.657376768',
               ...
               '2017-07-28 18:41:04.671513088',
               '2017-07-28 18:41:07.338664960',
               '2017-07-28 18:41:10.005816832',
               '2017-07-28 18:41:12.671969792',
               '2017-07-28 18:41:15.320121856',
               '2017-07-28 18:41:17.976272896',
               '2017-07-28 18:41:20.634423808',
               '2017-07-28 18:41:23.292577792',
               '2017-07-28 18:41:25.949729792',
               '2017-07-28 18:41:28.597880832'],
              dtype='datetime64[ns]', name='time1', length=529, freq=None))
```

--------------------------------

### Sv_offset Data Example

Source: https://echopype.readthedocs.io/en/latest/data-format-raw.html

Provides the Sv_offset values for each channel.

```python
array([0.7, 0.3, 0.3, 0.3])
```

--------------------------------

### Install echopype via Pip

Source: https://echopype.readthedocs.io/en/latest/_sources/installation.md

Use this command to install the latest release of echopype from PyPI.

```shell
pip install echopype
```

--------------------------------

### Simrad Transceiver Software Version Example

Source: https://echopype.readthedocs.io/en/latest/data-format-raw.html

Example of transceiver software versions for Simrad echosounders. This indicates the software version used by the transceiver.

```python
array(['2.20', '2.20', '2.20', '2.20', '2.20', '2.20'], dtype='<U4')
```

--------------------------------

### Sensor Offset Example

Source: https://echopype.readthedocs.io/en/latest/data-format-raw.html

Example of z-axis distance from the platform coordinate system origin to the sonar transducer.

```python
array([0., 0., 0., 0., 0., 0.])
```

--------------------------------

### Transceiver Software Version Data Example

Source: https://echopype.readthedocs.io/en/latest/data-format-raw.html

Represents the transceiver software version.

```python
array(['2.20', '2.20', '2.20'], dtype='<U4')
```

--------------------------------

### Create and Activate Conda Environment

Source: https://echopype.readthedocs.io/en/latest/contrib_setup.html

Create a Conda environment named 'echopype' using the conda-forge channel and install dependencies from requirement files. Activate the environment and install ipykernel for Jupyter integration.

```bash
# Create a conda environment using the supplied requirements files
# Note the last one docs/requirements.txt is only required for building docs
conda create -c conda-forge -n echopype --yes python=3.12 --file requirements.txt --file requirements-dev.txt --file docs/requirements.txt

# Switch to the newly built environment
conda activate echopype

# ipykernel is recommended, in order to use with JupyterLab and IPython for development
# We recommend you install JupyterLab separately
conda install -c conda-forge ipykernel

```

--------------------------------

### Tilt X Calibration Coefficient a Data Example

Source: https://echopype.readthedocs.io/en/latest/data-format-raw.html

The calibration coefficient a for Tilt-X.

```python
array(-45.77467916)
```

--------------------------------

### Number of Frequency Data Example

Source: https://echopype.readthedocs.io/en/latest/data-format-raw.html

Indicates the number of frequencies used.

```python
array(4)
```

--------------------------------

### Tilt X Calibration Coefficient d Data Example

Source: https://echopype.readthedocs.io/en/latest/data-format-raw.html

The calibration coefficient d for Tilt-X.

```python
array(-3.17633e-12)
```

--------------------------------

### AD Length Index Example

Source: https://echopype.readthedocs.io/en/latest/data-format-raw.html

Represents the AD data length index.

```python
PandasIndex(Index([0, 1], dtype='int64', name='ad_len'))
```

--------------------------------

### XML Burst Interval Data Example

Source: https://echopype.readthedocs.io/en/latest/data-format-raw.html

Specifies the time in seconds between bursts or pings.

```python
array(900)
```

--------------------------------

### Range Sample Data Example

Source: https://echopype.readthedocs.io/en/latest/data-format-raw.html

Represents the along-range sample number, starting from base 0.

```python
array([   0,    1,    2, ..., 6751, 6752, 6753])
```

--------------------------------

### Establish a Dask client

Source: https://echopype.readthedocs.io/en/latest/open-converted.html

Before combining data with `combine_echodata`, establish a Dask client. This example shows how to create a client with a local scheduler.

```python
from dask.distributed import Client
client = Client()  # create client with local scheduler

```

--------------------------------

### Clone Echopype Repository and Set Up Remotes

Source: https://echopype.readthedocs.io/en/latest/contrib_setup.html

Clone your fork of the Echopype repository and configure your local Git to recognize both your fork as 'origin' and the main OSOceanAcoustics repository as 'upstream'.

```bash
# Clone your fork
git clone https://github.com/YOUR_GITHUB_USERNAME/echopype.git

# Go into the cloned repo folder
cd echopype

# Add the OSOceanAcoustics repository as upstream
git remote add upstream https://github.com/OSOceanAcoustics/echopype.git

```

--------------------------------

### Time2 Index Example

Source: https://echopype.readthedocs.io/en/latest/data-format-raw.html

Example of PandasIndex for time2.

```python
PandasIndex(DatetimeIndex(['2018-12-13 09:46:00.747195904',
               '2018-12-13 09:46:01.457195008',
               '2018-12-13 09:46:01.930196992',
               '2018-12-13 09:46:03.337196032'],
              dtype='datetime64[ns]', name='time2', freq=None))
```

--------------------------------

### Time1 Index Example

Source: https://echopype.readthedocs.io/en/latest/data-format-raw.html

Example of PandasIndex for time1.

```python
PandasIndex(DatetimeIndex(['2018-12-13 09:46:01.363802112',
               '2018-12-13 09:46:01.363802112',
               '2018-12-13 09:46:01.363802112',
               '2018-12-13 09:46:02.363802112',
               '2018-12-13 09:46:02.363802112',
               '2018-12-13 09:46:02.363802112',
               '2018-12-13 09:46:03.363802112',
               '2018-12-13 09:46:03.363802112',
               '2018-12-13 09:46:03.363802112',
               '2018-12-13 09:46:04.363802112',
               '2018-12-13 09:46:04.363802112',
               '2018-12-13 09:46:04.363802112'],
              dtype='datetime64[ns]', name='time1', freq=None))
```

--------------------------------

### Configuration XML Example

Source: https://echopype.readthedocs.io/en/latest/data-format-raw.html

Shows the configuration XML content as a string. This parameter holds the XML configuration for the sonar system.

```xml
<?xml version="1.0" encoding="u..."
```

--------------------------------

### Clone Repository and Set Up Remotes

Source: https://echopype.readthedocs.io/en/latest/_sources/contrib_setup.md

Clone your Echopype fork and configure git remotes to point to your fork as 'origin' and the main repository as 'upstream'.

```shell
git clone https://github.com/YOUR_GITHUB_USERNAME/echopype.git

cd echopype

git remote add upstream https://github.com/OSOceanAcoustics/echopype.git
```

--------------------------------

### Nominal Frequency Example

Source: https://echopype.readthedocs.io/en/latest/data-format-raw.html

Example of transducer frequency in Hz.

```python
array([ 18000., 120000.,  70000.,  38000., 333000., 200000.])
```

--------------------------------

### NMEA Dataset Example

Source: https://echopype.readthedocs.io/en/latest/data-format-raw.html

Example of an xarray.Dataset containing NMEA datagrams.

```python
<xarray.Dataset>
Dimensions:        (time1: 142)
Coordinates:
  * time1          (time1) datetime64[ns] 2018-12-13T09:46:00.747195904 ... 2...
Data variables:
    NMEA_datagram  (time1) <U93 '$SDVLW,2559.924,N,2559.924,N' ... '$PBFG,1,0...>
Attributes:
    description:  All NMEA sensor datagrams
```

--------------------------------

### Deploy Docker Services for Testing

Source: https://echopype.readthedocs.io/en/latest/contrib_setup.html

Install and deploy the necessary Docker containers for Echopype testing. This command also handles the download of required test data files.

```bash
# Install and/or deploy the echopype docker containers for testing.
# Test data files will be downloaded
python .ci_helpers/docker/setup-services.py --deploy

```

--------------------------------

### Build Echopype Documentation Locally

Source: https://echopype.readthedocs.io/en/latest/_sources/contrib_setup.md

Build the project's documentation using Jupyter Book. This command generates the HTML files for local viewing.

```shell
jupyter-book build docs/source --path-output docs
```

--------------------------------

### Transmit Frequency Start

Source: https://echopype.readthedocs.io/en/latest/data-format-raw.html

The starting frequency of the transmitted pulse in Hertz. Must be non-negative.

```python
array([38000., 18000.])
```

--------------------------------

### EchoData Platform Example

Source: https://echopype.readthedocs.io/en/latest/data-format-raw.html

Illustrates the platform information in an EchoData object, containing details like latitude, longitude, and time series data.

```python
<xarray.Dataset>
Dimensions:                      (time1: 12, time2: 4, channel: 6)
Coordinates:
  * channel                      (channel) <U22 'WBT 714581-15 ES18' ... 'WBT... 
  * time1                        (time1) datetime64[ns] 2018-12-13T09:46:01.3... 
  * time2                        (time2) datetime64[ns] 2018-12-13T09:46:00.7... 
Data variables: (12/23) 
    latitude                     (time1) float64 60.41 60.41 ... 60.41 60.41
    longitude                    (time1) float64 5.312 5.312 ... 5.312 5.312
```

--------------------------------

### NMEA Datagram Timestamps Example

Source: https://echopype.readthedocs.io/en/latest/data-format-raw.html

Example of time coordinate corresponding to NMEA sensor data.

```python
array(['2018-12-13T09:46:00.747195904', '2018-12-13T09:46:01.126800896',
       '2018-12-13T09:46:01.363802112', '2018-12-13T09:46:01.363802112',
       '2018-12-13T09:46:01.363802112', '2018-12-13T09:46:01.363802112',
       '2018-12-13T09:46:01.363802112', '2018-12-13T09:46:01.363802112',
       '2018-12-13T09:46:01.363802112', '2018-12-13T09:46:01.363802112',
       '2018-12-13T09:46:01.363802112', '2018-12-13T09:46:01.363802112',
       '2018-12-13T09:46:01.363802112', '2018-12-13T09:46:01.363802112',
       '2018-12-13T09:46:01.363802112', '2018-12-13T09:46:01.363802112'])
```

--------------------------------

### Deploy Docker Services for Testing

Source: https://echopype.readthedocs.io/en/latest/_sources/contrib_setup.md

Installs and deploys the necessary Docker containers for Echopype testing, including downloading test data files.

```shell
python .ci_helpers/docker/setup-services.py --deploy
```

--------------------------------

### Install echopype via Conda

Source: https://echopype.readthedocs.io/en/latest/_sources/installation.md

Use this command to install the latest release of echopype from the conda-forge channel.

```shell
conda install -c conda-forge echopype
```

--------------------------------

### Access Raw File from HTTP Server

Source: https://echopype.readthedocs.io/en/latest/convert.html

This example demonstrates how to open a raw file located on a web server using its URL. The `fsspec` package enables this functionality.

```python
ed = open_raw(
    "https://mydomain.com/my/dir/D20170615-T190214.raw",  # file on http server
    sonar_model="EK80"
)
```

--------------------------------

### Create Conda Environment and Install Dependencies

Source: https://echopype.readthedocs.io/en/latest/_sources/contrib_setup.md

Create a conda environment named 'echopype' using conda-forge, installing Python and project dependencies from requirement files. Includes ipykernel for Jupyter development and installs echopype in editable mode with plotting extras.

```shell
conda create -c conda-forge -n echopype --yes python=3.12 --file requirements.txt --file requirements-dev.txt --file docs/requirements.txt

conda activate echopype

conda install -c conda-forge ipykernel

pip install -e ".[plot]"
```

--------------------------------

### Enable Pooch Test Data Download (Linux/macOS)

Source: https://echopype.readthedocs.io/en/latest/_sources/contrib_setup.md

Set the USE_POOCH environment variable to 'True' on Linux or macOS to enable automatic downloading of test data into the local Pooch cache.

```shell
export USE_POOCH=True
```

--------------------------------

### Install echopype via Conda

Source: https://echopype.readthedocs.io/en/latest/installation.html

Installs the latest release of echopype using the conda-forge channel. Recommended for managing dependencies.

```bash
# Install via conda-forge
$ conda install -c conda-forge echopype

```

--------------------------------

### Transmit Frequency Start

Source: https://echopype.readthedocs.io/en/latest/data-format-raw.html

Specifies the start frequency of the transmitted pulse in Hertz. This value is a float64 and has a minimum valid value of 0.0.

```python
array([[165000., 165000., 165000., 165000., 165000., 165000., 165000.,
        165000.],
       [ 45000.,  45000.,  45000.,  45000.,  45000.,  45000.,  45000.,
         45000.],
       [ 90000.,  90000.,  90000.,  90000.,  90000.,  90000.,  90000.,
         90000.]])
```

--------------------------------

### EchoData Environment Example

Source: https://echopype.readthedocs.io/en/latest/data-format-raw.html

Shows the environmental data within an EchoData object, including depth, temperature, salinity, and sound speed profiles.

```python
<xarray.Dataset>
Dimensions:                       (time1: 1, sound_velocity_profile_depth: 2)
Coordinates:
  * time1                         (time1) datetime64[ns] 2018-12-13T09:46:00....
  * sound_velocity_profile_depth  (sound_velocity_profile_depth) float64 1.0 ...
Data variables:
    depth                         (time1) float64 50.0
    acidity                       (time1) float64 8.0
    salinity                      (time1) float64 32.8
    temperature                   (time1) float64 9.8
    sound_speed_indicative        (time1) float64 1.48e+03
    sound_velocity_profile        (time1, sound_velocity_profile_depth) float64 ...
    sound_velocity_source         (time1) <U6 'Manual'
    transducer_name               (time1) <U7 'Unknown'
    transducer_sound_speed        (time1) float64 1.49e+03
```

--------------------------------

### Enable Pooch for Test Data Download (Linux/macOS)

Source: https://echopype.readthedocs.io/en/latest/contrib_setup.html

Set the USE_POOCH environment variable to 'True' on Linux or macOS to enable the automatic download of test data into the local Pooch cache when tests are run.

```bash
export USE_POOCH=True

```

--------------------------------

### Install echopype via Pip

Source: https://echopype.readthedocs.io/en/latest/installation.html

Installs the latest release of echopype using pip. Ensure you have a compatible Python version (3.10-3.12).

```bash
# Install via pip
$ pip install echopype

```

--------------------------------

### Example Data Array Structure

Source: https://echopype.readthedocs.io/en/latest/data-format-changes.html

Illustrates the multi-dimensional array structure for sonar data, showing dimensions like frequency, ping time, and range bin.

```text
[[  -2.,   -3.,   -3., ...,  107.,  -32.,  -20.],
        [  -2.,   -3.,   -3., ...,  -13.,  -99.,  -39.],
        [  -2.,   -3.,   -3., ...,   19.,   54.,   33.]]])
```

--------------------------------

### Enable Pooch Test Data Download (Windows PowerShell)

Source: https://echopype.readthedocs.io/en/latest/_sources/contrib_setup.md

Set the USE_POOCH environment variable to 'True' in Windows PowerShell to enable automatic downloading of test data into the local Pooch cache.

```powershell
$env:USE_POOCH="True"
```

--------------------------------

### Simrad Backscatter Raw Imaginary Part Example

Source: https://echopype.readthedocs.io/en/latest/data-format-raw.html

Example of the imaginary part of raw backscatter measurements for Simrad echosounders. Units are in decibels (dB).

```python
array([[[[-9.19969473e-03,  3.31364311e-02,  7.40730437e-03,
           9.50816199e-02],
         [ 7.51427514e-03, -3.74089368e-02,  1.86177538e-04,
          -8.94665197e-02],
         [-5.42381266e-03,  3.55778970e-02, -9.93139949e-03,
           7.26207718e-02],
         [            nan,             nan,             nan,
                      nan],
         [            nan,             nan,             nan,
                      nan],
         [            nan,             nan,             nan,
                      nan]],

        [[ 3.10946498e-02,  9.44385398e-03,  7.16440156e-02,
           3.82631160e-02],
         [-3.52116413e-02, -1.60190258e-02, -6.89585283e-02,
          -3.76524962e-02],
         [ 3.65540572e-02,  2.19393391e-02,  5.77947013e-02,
           3.24024446e-02],
         [            nan,             nan,             nan,
                      nan],
         [            nan,             nan,             nan,
                      nan],
         [            nan,             nan,             nan,
                      nan]],

        [[-6.38506981e-03, -3.51283536e-03, -1.64585176e-03,
           7.27002183e-03],
         [ 7.74312718e-03, -4.35542269e-03,  1.10611236e-02,
           1.69340242e-02],
         [-5.95780322e-03,  4.52320790e-03, -1.12746796e-02,
          -3.92384566e-02],
         [            nan,             nan,             nan,
                      nan],
         [            nan,             nan,             nan,
                      nan],
         [            nan,             nan,             nan,
                      nan]]]], dtype=float32)
```

--------------------------------

### Basic Seafloor Detection

Source: https://echopype.readthedocs.io/en/latest/api/echopype.mask.detect_seafloor.html

Example of using the 'basic' method for seafloor detection. Requires specifying the Sv variable name, channel, threshold, and optional parameters like offset and bins to skip from the surface.

```python
detect_seafloor(ds, "basic", {
    "var_name": "Sv", "channel": "GPT  38 kHz ...",
    "threshold": -50, "offset_m": 0.5, "bin_skip_from_surface": 200
})
```

--------------------------------

### Assemble List of EchoData Objects for Combination

Source: https://echopype.readthedocs.io/en/latest/_sources/open-converted.ipynb

Create a list of EchoData objects by opening converted files (Zarr in this example). These files are lazily loaded, allowing for efficient combination of large datasets.

```python
ed_list = []
for converted_file in ["convertedfile1.zarr", "convertedfile2.zarr"]:
    ed_list.append(ep.open_converted(converted_file))  # already converted files are lazy-loaded
```

--------------------------------

### Simrad Backscatter Raw Real Part Example

Source: https://echopype.readthedocs.io/en/latest/data-format-raw.html

Example of the real part of raw backscatter measurements for Simrad echosounders. Units are in decibels (dB).

```python
array([[[[-1.28920395e-02,  3.42344306e-02,  1.46622611e-02,
          -5.56103513e-03],
         [ 1.44484267e-02, -3.77746075e-02, -1.46317678e-02,
           6.24768576e-03],
         [-1.44178672e-02,  3.74084488e-02,  1.30753778e-02,
          -8.68065935e-03],
         [            nan,             nan,             nan,
                      nan],
         [            nan,             nan,             nan,
                      nan],
         [            nan,             nan,             nan,
                      nan]],

        [[ 9.77926049e-03, -1.55777158e-02, -3.41123864e-02,
           4.40140208e-03],
         [-1.32277710e-02,  2.13288255e-02,  3.44788209e-02,
          -8.65021069e-03],
         [ 1.66289546e-02, -2.52961051e-02, -3.02399881e-02,
           1.39907384e-02],
         [            nan,             nan,             nan,
                      nan],
         [            nan,             nan,             nan,
                      nan],
         [            nan,             nan,             nan,
                      nan]],

        [[-1.29483407e-03, -1.70557443e-02,  2.13893913e-02,
           3.81367430e-02],
         [ 5.51529787e-03,  1.41125275e-02, -1.61404330e-02,
          -4.29014713e-02],
         [-1.03288367e-02, -8.40559881e-03,  8.22297111e-03,
           2.92023439e-02],
         [            nan,             nan,             nan,
                      nan],
         [            nan,             nan,             nan,
                      nan],
         [            nan,             nan,             nan,
                      nan]]]], dtype=float32)
```

--------------------------------

### Compute Frequency-Differencing Mask

Source: https://echopype.readthedocs.io/en/latest/api/echopype.mask.frequency_differencing.html

Demonstrates how to compute a frequency-differencing mask using a mock Dataset and channel selection. This example requires creating mock Sv and frequency_nominal data, constructing a Dataset, and then calling the frequency_differencing function with specific channel criteria.

```python
import echopype
import xarray as xr
import numpy as np

n = 5 # set the number of ping times and range samples

# create mock Sv data
Sv_da = xr.DataArray(data=np.stack([np.arange(n**2).reshape(n,n), np.identity(n)]),
                     coords={"channel": ['chan1', 'chan2'],
                             "ping_time": np.arange(n), "range_sample":np.arange(n)})

# obtain mock frequency_nominal data
freq_nom = xr.DataArray(data=np.array([1.0, 2.0]),
                        coords={"channel": ['chan1', 'chan2']})

# construct mock Sv Dataset
Sv_ds = xr.Dataset(data_vars={"Sv": Sv_da, "frequency_nominal": freq_nom})

# compute frequency-differencing mask using channel names
echopype.mask.frequency_differencing(source_Sv=Sv_ds, storage_options={}, 
                                     freqABEq=None, chanABEq = '"chan1" - "chan2">=10.0dB')

```

--------------------------------

### Simrad Equivalent Beam Angle Example

Source: https://echopype.readthedocs.io/en/latest/data-format-raw.html

Example of equivalent beam angle data for Simrad echosounders. This parameter represents the major angle and is measured in steradians.

```python
array([-17. , -20.7, -20.7, -20.7, -20.7, -20.7])
```

--------------------------------

### Kongsberg Simrad EK60 Processed Roll Data Example

Source: https://echopype.readthedocs.io/en/latest/data-format-raw.html

This snippet shows an example of processed 'roll' data, presented as a NumPy array. It includes the data type and a sample of the values.

```python
array([-3.74009997e-01, -7.12723017e-01, -4.91404176e-01, -6.71638310e-01,
       -1.45054793e+00,  6.84964880e-02, -6.74356401e-01, -6.14172876e-01,
       -1.28917551e+00, -7.91940212e-01, -8.99480820e-01, -1.03791916e+00,
       -2.89999992e-01, -1.66740417e+00, -4.76876676e-01, -2.00921917e+00,
       -2.17319161e-01, -1.25999999e+00, -1.51003465e-01, -1.47203827e+00,
       -2.38418192e-01, -6.55282080e-01, -4.67003137e-01, -1.49277246e+00,
        8.38683397e-02, -1.85859585e+00, -4.39999998e-01, -4.93865788e-01,
       -1.39167666e+00, -4.62057531e-01, -1.32833862e+00, -2.27255791e-01,
       -1.40832829e+00,  3.06862980e-01, -1.88089371e+00, -6.67472854e-02,
       -7.77107999e-02, -8.88284445e-01, -1.79062247e-01, -7.09999979e-01,
       -2.19676644e-01, -5.75265825e-01, -2.43119165e-01, -6.58352315e-01,
       -1.51162967e-01, -9.43595827e-01, -8.89904201e-01, -6.48876667e-01,
       -1.46838337e-01, -5.44961691e-01, -1.14999998e+00, -8.12476635e-01,
       -8.25808287e-01, -6.51030004e-01, -3.77565861e-01, -1.53999996e+00,
       -3.64357501e-01, -7.97468364e-01, -6.70000017e-01, -1.09000003e+00,
       -5.62633514e-01, -9.41298544e-01, -3.55338812e-01, -2.04030693e-01,
       -1.04959583e+00, -3.41880828e-01, -5.89999974e-01, -5.56876659e-01,
       -6.37184799e-01, -1.08106852e+00, -3.01977936e-02, -9.35161650e-01,
       -9.28095877e-01, -9.14577246e-01, -7.83651531e-01, -7.66196787e-01,
       -8.11438322e-01, -1.12000000e+00, -2.98904181e-01, -9.51719165e-01,
       -2.12419167e-01, -6.16719186e-01, -5.32519162e-01,  1.82180837e-01,
       -7.97080815e-01, -4.23361659e-01, -6.39038324e-01,  3.30000013e-01,
       -4.93761659e-01, -8.03276658e-01, -4.29918505e-02, -6.89999998e-01,
       -1.67072445e-01, -3.24070692e-01, -6.02847755e-01, -4.03961658e-01,
       -5.29999971e-01,  1.28619164e-01, -7.20000029e-01, -9.48242486e-01,
       -2.30861261e-01, -3.71793896e-01, -4.21877354e-01, -3.06770980e-01,
       -4.12488341e-01, -2.91785359e-01, -4.22654301e-01, -7.25975275e-01])
```
Tokens:
Raw