Benchmark multi-model/multi-view models.

Note

This page is a reference documentation. It only explains the class signature, and not how to use it. Please refer to the gallery for the big picture.

class mmbench.dataset.EUAIMSDataset(root, train=True, transform=None, flatten=False, seed=42, scaler=None)[source]¶

From the EUAIMS cohort the target and background datasets are composed of T1w MRI FreeSurfer ROI features of ASD patients and TD controls, respectively.

__init__(root, train=True, transform=None, flatten=False, seed=42, scaler=None)[source]¶

Init class.

Parameters

root : str

root directory of dataset where data will be saved.

train : bool, default True

specifies training or test dataset.

transform : callable, default None

optional transform to be applied on a sample.

flatten : bool, default False

optionally select all data.

seed : int, default 42

for reproducibility fix a seed.

scaler : sklearn-like scaler, default None

optionally set a fitted scaler.

get_data()[source]¶

Get the background and target data.

Returns

background : array (N, n_channels, *)

the background data.

background_labels : array (N, )

the background labels.

target : array (M, n_channels, *)

the target data.

target_labels : array (M, )

the target labels.

is_file(path)[source]¶: Check wethe a EUAIMS data resource file is here.