Note

Go to the end to download the full example code. or to run this example in your browser via Binder

Voxel-Based Morphometry on Oasis dataset with Space-Net prior

Predicting age from gray-matter concentration maps from OASIS dataset. Note that age is a continuous variable, we use the regressor here, and not the classification object.

See also

The documentation: SpaceNet: decoding with spatial structure for better maps.

For more information see the dataset description.

Load the Oasis VBM dataset

import numpy as np

from nilearn import datasets

n_subjects = 200
dataset_files = datasets.fetch_oasis_vbm(n_subjects=n_subjects)
age = dataset_files.ext_vars["age"].astype(float)
age = np.array(age)
gm_imgs = np.array(dataset_files.gray_matter_maps)


# Split data into training set and test set
from sklearn.model_selection import train_test_split
from sklearn.utils import check_random_state

rng = check_random_state(42)
gm_imgs_train, gm_imgs_test, age_train, age_test = train_test_split(
    gm_imgs, age, train_size=0.6, random_state=rng
)

# Sort test data for better visualization (trend, etc.)
perm = np.argsort(age_test)[::-1]
age_test = age_test[perm]
gm_imgs_test = gm_imgs_test[perm]

[fetch_oasis_vbm] Dataset found in /home/runner/nilearn_data/oasis1

Fit the SpaceNet and predict with it

To save time (because these are anat images with many voxels), we include only the 5-percent voxels most correlated with the age variable to fit. Also, we set memory_level=2 so that more of the intermediate computations are cached. We used a graph-net penalty here but more beautiful results can be obtained using the TV-l1 penalty, at the expense of longer runtimes. Also, you may pass and n_jobs=<some_high_value> to the SpaceNetRegressor class, to take advantage of a multi-core system.

from nilearn.decoding import SpaceNetRegressor

decoder = SpaceNetRegressor(
    memory="nilearn_cache",
    penalty="graph-net",
    screening_percentile=5.0,
    memory_level=2,
    standardize="zscore_sample",
    n_jobs=2,
)
decoder.fit(gm_imgs_train, age_train)  # fit
coef_img = decoder.coef_img_
y_pred = decoder.predict(gm_imgs_test).ravel()  # predict
mse = np.mean(np.abs(age_test - y_pred))
print(f"Mean square error (MSE) on the predicted age: {mse:.2f}")

[SpaceNetRegressor.fit] Loading data from
array(['/home/runner/nilearn_data/oasis1/OAS1_0003_MR1/mwrc1OAS1_0003_MR1_mpr_an
on_fslswapdim_bet.nii.gz',
       '/home/runner/nilearn_data/oasis1/OAS1_0086_MR1/mwrc1OAS1_0086_MR1_mpr_an
on_fslswapdim_bet.nii.gz',
       '/home/runner/nilearn_data/oasis1/OAS1_0052_MR1/mwrc1OAS1_0052_MR1_mpr_an
on_fslswapdim_bet.nii.gz',
       '/home/runner/nilearn_data/oasis1/OAS1_0211_MR1/mwrc1OAS1_0211_MR1_mpr_an
on_fslswapdim_bet.nii.gz',
       '/home/runner/nilearn_data/oasis1/OAS1_0216_MR1/mwrc1OAS1_0216_MR1_mpr_an
on_fslswapdim_bet.nii.gz',
       '/home/runner/nilearn_data/oasis1/OAS1_0096_MR1/mwrc1OAS1_0096_MR1_mpr_an
on_fslswapdim_bet.nii.gz',
       '/home/runner/nilearn_data/oasis1/OAS1_0182_MR1/mwrc1OAS1_0182_MR1_mpr_an
on_fslswapdim_bet.nii.gz',
       '/home/runner/nilearn_data/oasis1/OAS1_0041_MR1/mwrc1OAS1_0041_MR1_mpr_an
on_fslswapdim_bet.nii.gz',
       '/home/runner/nilearn_data/oasis1/OAS1_0214_MR1/mwrc1OAS1_0214_MR1_mpr_an
on_fslswapdim_bet.nii.gz',
       '/home/runner/nilearn_data/oasis1/OAS1_0069_MR1/mwrc1OAS1_0069_MR1_mpr_an
on_fslswapdim_bet.nii.gz',
       '/home/runner/nilearn_data/oasis1/OAS1_0025_MR1/mwrc1OAS1_0025_MR1_mpr_an
on_fslswapdim_bet.nii.gz',
       '/home/runner/nilearn_data/oasis1/OAS1_0159_MR1/mwrc1OAS1_0159_MR1_mpr_an
on_fslswapdim_bet.nii.gz',
       '/home/runner/nilearn_data/oasis1/OAS1_0113_MR1/mwrc1OAS1_0113_MR1_mpr_an
on_fslswapdim_bet.nii.gz',
       '/home/runner/nilearn_data/oasis1/OAS1_0038_MR1/mwrc1OAS1_0038_MR1_mpr_an
on_fslswapdim_bet.nii.gz',
       '/home/runner/nilearn_data/oasis1/OAS1_0013_MR1/mwrc1OAS1_0013_MR1_mpr_an
on_fslswapdim_bet.nii.gz',
       '/home/runner/nilearn_data/oasis1/OAS1_0220_MR1/mwrc1OAS1_0220_MR1_mpr_an
on_fslswapdim_bet.nii.gz',
       '/home/runner/nilearn_data/oasis1/OAS1_0180_MR1/mwrc1OAS1_0180_MR1_mpr_an
on_fslswapdim_bet.nii.gz',
       '/home/runner/nilearn_data/oasis1/OAS1_0007_MR1/mwrc1OAS1_0007_MR1_mpr_an
on_fslswapdim_bet.nii.gz',
       '/home/runner/nilearn_data/oasis1/OAS1_0031_MR1/mwrc1OAS1_0031_MR1_mpr_an
on_fslswapdim_bet.nii.gz',
       '/home/runner/nilearn_data/oasis1/OAS1_0135_MR1/mwrc1OAS1_0135_MR1_mpr_an
on_fslswapdim_bet.nii.gz',
       '/home/runner/nilearn_data/oasis1/OAS1_0005_MR1/mwrc1OAS1_0005_MR1_mpr_an
on_fslswapdim_bet.nii.gz',
       '/home/runner/nilearn_data/oasis1/OAS1_0037_MR1/mwrc1OAS1_0037_MR1_mpr_an
on_fslswapdim_bet.nii.gz',
       '/home/runner/nilearn_data/oasis1/OAS1_0160_MR1/mwrc1OAS1_0160_MR1_mpr_an
on_fslswapdim_bet.nii.gz',
       '/home/runner/nilearn_data/oasis1/OAS1_0163_MR1/mwrc1OAS1_0163_MR1_mpr_an
on_fslswapdim_bet.nii.gz',
       '/home/runner/nilearn_data/oasis1/OAS1_0123_MR1/mwrc1OAS1_0123_MR1_mpr_an
on_fslswapdim_bet.nii.gz',
       '/home/runner/nilearn_data/oasis1/OAS1_0162_MR1/mwrc1OAS1_0162_MR1_mpr_an
on_fslswapdim_bet.nii.gz',
       '/home/runner/nilearn_data/oasis1/OAS1_0012_MR1/mwrc1OAS1_0012_MR1_mpr_an
on_fslswapdim_bet.nii.gz',
       '/home/runner/nilearn_data/oasis1/OAS1_0070_MR1/mwrc1OAS1_0070_MR1_mpr_an
on_fslswapdim_bet.nii.gz',
       '/home/runner/nilearn_data/oasis1/OAS1_0126_MR1/mwrc1OAS1_0126_MR1_mpr_an
on_fslswapdim_bet.nii.gz',
       '/home/runner/nilearn_data/oasis1/OAS1_0164_MR1/mwrc1OAS1_0164_MR1_mpr_an
on_fslswapdim_bet.nii.gz',
       '/home/runner/nilearn_data/oasis1/OAS1_0188_MR1/mwrc1OAS1_0188_MR1_mpr_an
on_fslswapdim_bet.nii.gz',
       '/home/runner/nilearn_data/oasis1/OAS1_0001_MR1/mwrc1OAS1_0001_MR1_mpr_an
on_fslswapdim_bet.nii.gz',
       '/home/runner/nilearn_data/oasis1/OAS1_0226_MR1/mwrc1OAS1_0226_MR1_mpr_an
on_fslswapdim_bet.nii.gz',
       '/home/runner/nilearn_data/oasis1/OAS1_0173_MR1/mwrc1OAS1_0173_MR1_mpr_an
on_fslswapdim_bet.nii.gz',
       '/home/runner/nilearn_data/oasis1/OAS1_0078_MR1/mwrc1OAS1_0078_MR1_mpr_an
on_fslswapdim_bet.nii.gz',
       '/home/runner/nilearn_data/oasis1/OAS1_0138_MR1/mwrc1OAS1_0138_MR1_mpr_an
on_fslswapdim_bet.nii.gz',
       '/home/runner/nilearn_data/oasis1/OAS1_0072_MR1/mwrc1OAS1_0072_MR1_mpr_an
on_fslswapdim_bet.nii.gz',
       '/home/runner/nilearn_data/oasis1/OAS1_0050_MR1/mwrc1OAS1_0050_MR1_mpr_an
on_fslswapdim_bet.nii.gz',
       '/home/runner/nilearn_data/oasis1/OAS1_0184_MR1/mwrc1OAS1_0184_MR1_mpr_an
on_fslswapdim_bet.nii.gz',
       '/home/runner/nilearn_data/oasis1/OAS1_0032_MR1/mwrc1OAS1_0032_MR1_mpr_an
on_fslswapdim_bet.nii.gz',
       '/home/runner/nilearn_data/oasis1/OAS1_0045_MR1/mwrc1OAS1_0045_MR1_mpr_an
on_fslswapdim_bet.nii.gz',
       '/home/runner/nilearn_data/oasis1/OAS1_0121_MR1/mwrc1OAS1_0121_MR1_mpr_an
on_fslswapdim_bet.nii.gz',
       '/home/runner/nilearn_data/oasis1/OAS1_0176_MR1/mwrc1OAS1_0176_MR1_mpr_an
on_fslswapdim_bet.nii.gz',
       '/home/runner/nilearn_data/oasis1/OAS1_0177_MR1/mwrc1OAS1_0177_MR1_mpr_an
on_fslswapdim_bet.nii.gz',
       '/home/runner/nilearn_data/oasis1/OAS1_0029_MR1/mwrc1OAS1_0029_MR1_mpr_an
on_fslswapdim_bet.nii.gz',
       '/home/runner/nilearn_data/oasis1/OAS1_0026_MR1/mwrc1OAS1_0026_MR1_mpr_an
on_fslswapdim_bet.nii.gz',
       '/home/runner/nilearn_data/oasis1/OAS1_0208_MR1/mwrc1OAS1_0208_MR1_mpr_an
on_fslswapdim_bet.nii.gz',
       '/home/runner/nilearn_data/oasis1/OAS1_0165_MR1/mwrc1OAS1_0165_MR1_mpr_an
on_fslswapdim_bet.nii.gz',
       '/home/runner/nilearn_data/oasis1/OAS1_0091_MR1/mwrc1OAS1_0091_MR1_mpr_an
on_fslswapdim_bet.nii.gz',
       '/home/runner/nilearn_data/oasis1/OAS1_0044_MR1/mwrc1OAS1_0044_MR1_mpr_an
on_fslswapdim_bet.nii.gz',
       '/home/runner/nilearn_data/oasis1/OAS1_0190_MR1/mwrc1OAS1_0190_MR1_mpr_an
on_fslswapdim_bet.nii.gz',
       '/home/runner/nilearn_data/oasis1/OAS1_0053_MR1/mwrc1OAS1_0053_MR1_mpr_an
on_fslswapdim_bet.nii.gz',
       '/home/runner/nilearn_data/oasis1/OAS1_0106_MR1/mwrc1OAS1_0106_MR1_mpr_an
on_fslswapdim_bet.nii.gz',
       '/home/runner/nilearn_data/oasis1/OAS1_0174_MR1/mwrc1OAS1_0174_MR1_mpr_an
on_fslswapdim_bet.nii.gz',
       '/home/runner/nilearn_data/oasis1/OAS1_0049_MR1/mwrc1OAS1_0049_MR1_mpr_an
on_fslswapdim_bet.nii.gz',
       '/home/runner/nilearn_data/oasis1/OAS1_0155_MR1/mwrc1OAS1_0155_MR1_mpr_an
on_fslswapdim_bet.nii.gz',
       '/home/runner/nilearn_data/oasis1/OAS1_0004_MR1/mwrc1OAS1_0004_MR1_mpr_an
on_fslswapdim_bet.nii.gz',
       '/home/runner/nilearn_data/oasis1/OAS1_0117_MR1/mwrc1OAS1_0117_MR1_mpr_an
on_fslswapdim_bet.nii.gz',
       '/home/runner/nilearn_data/oasis1/OAS1_0060_MR1/mwrc1OAS1_0060_MR1_mpr_an
on_fslswapdim_bet.nii.gz',
       '/home/runner/nilearn_data/oasis1/OAS1_0148_MR1/mwrc1OAS1_0148_MR1_mpr_an
on_fslswapdim_bet.nii.gz',
       '/home/runner/nilearn_data/oasis1/OAS1_0204_MR1/mwrc1OAS1_0204_MR1_mpr_an
on_fslswapdim_bet.nii.gz',
       '/home/runner/nilearn_data/oasis1/OAS1_0202_MR1/mwrc1OAS1_0202_MR1_mpr_an
on_fslswapdim_bet.nii.gz',
       '/home/runner/nilearn_data/oasis1/OAS1_0209_MR1/mwrc1OAS1_0209_MR1_mpr_an
on_fslswapdim_bet.nii.gz',
       '/home/runner/nilearn_data/oasis1/OAS1_0055_MR1/mwrc1OAS1_0055_MR1_mpr_an
on_fslswapdim_bet.nii.gz',
       '/home/runner/nilearn_data/oasis1/OAS1_0090_MR1/mwrc1OAS1_0090_MR1_mpr_an
on_fslswapdim_bet.nii.gz',
       '/home/runner/nilearn_data/oasis1/OAS1_0039_MR1/mwrc1OAS1_0039_MR1_mpr_an
on_fslswapdim_bet.nii.gz',
       '/home/runner/nilearn_data/oasis1/OAS1_0009_MR1/mwrc1OAS1_0009_MR1_mpr_an
on_fslswapdim_bet.nii.gz',
       '/home/runner/nilearn_data/oasis1/OAS1_0124_MR1/mwrc1OAS1_0124_MR1_mpr_an
on_fslswapdim_bet.nii.gz',
       '/home/runner/nilearn_data/oasis1/OAS1_0103_MR1/mwrc1OAS1_0103_MR1_mpr_an
on_fslswapdim_bet.nii.gz',
       '/home/runner/nilearn_data/oasis1/OAS1_0094_MR1/mwrc1OAS1_0094_MR1_mpr_an
on_fslswapdim_bet.nii.gz',
       '/home/runner/nilearn_data/oasis1/OAS1_0200_MR1/mwrc1OAS1_0200_MR1_mpr_an
on_fslswapdim_bet.nii.gz',
       '/home/runner/nilearn_data/oasis1/OAS1_0205_MR1/mwrc1OAS1_0205_MR1_mpr_an
on_fslswapdim_bet.nii.gz',
       '/home/runner/nilearn_data/oasis1/OAS1_0101_MR1/mwrc1OAS1_0101_MR1_mpr_an
on_fslswapdim_bet.nii.gz',
       '/home/runner/nilearn_data/oasis1/OAS1_0010_MR1/mwrc1OAS1_0010_MR1_mpr_an
on_fslswapdim_bet.nii.gz',
       '/home/runner/nilearn_data/oasis1/OAS1_0015_MR1/mwrc1OAS1_0015_MR1_mpr_an
on_fslswapdim_bet.nii.gz',
       '/home/runner/nilearn_data/oasis1/OAS1_0067_MR1/mwrc1OAS1_0067_MR1_mpr_an
on_fslswapdim_bet.nii.gz',
       '/home/runner/nilearn_data/oasis1/OAS1_0193_MR1/mwrc1OAS1_0193_MR1_mpr_an
on_fslswapdim_bet.nii.gz',
       '/home/runner/nilearn_data/oasis1/OAS1_0146_MR1/mwrc1OAS1_0146_MR1_mpr_an
on_fslswapdim_bet.nii.gz',
       '/home/runner/nilearn_data/oasis1/OAS1_0019_MR1/mwrc1OAS1_0019_MR1_mpr_an
on_fslswapdim_bet.nii.gz',
       '/home/runner/nilearn_data/oasis1/OAS1_0080_MR1/mwrc1OAS1_0080_MR1_mpr_an
on_fslswapdim_bet.nii.gz',
       '/home/runner/nilearn_data/oasis1/OAS1_0199_MR1/mwrc1OAS1_0199_MR1_mpr_an
on_fslswapdim_bet.nii.gz',
       '/home/runner/nilearn_data/oasis1/OAS1_0150_MR1/mwrc1OAS1_0150_MR1_mpr_an
on_fslswapdim_bet.nii.gz',
       '/home/runner/nilearn_data/oasis1/OAS1_0189_MR1/mwrc1OAS1_0189_MR1_mpr_an
on_fslswapdim_bet.nii.gz',
       '/home/runner/nilearn_data/oasis1/OAS1_0207_MR1/mwrc1OAS1_0207_MR1_mpr_an
on_fslswapdim_bet.nii.gz',
       '/home/runner/nilearn_data/oasis1/OAS1_0071_MR1/mwrc1OAS1_0071_MR1_mpr_an
on_fslswapdim_bet.nii.gz',
       '/home/runner/nilearn_data/oasis1/OAS1_0061_MR1/mwrc1OAS1_0061_MR1_mpr_an
on_fslswapdim_bet.nii.gz',
       '/home/runner/nilearn_data/oasis1/OAS1_0120_MR1/mwrc1OAS1_0120_MR1_mpr_an
on_fslswapdim_bet.nii.gz',
       '/home/runner/nilearn_data/oasis1/OAS1_0056_MR1/mwrc1OAS1_0056_MR1_mpr_an
on_fslswapdim_bet.nii.gz',
       '/home/runner/nilearn_data/oasis1/OAS1_0223_MR1/mwrc1OAS1_0223_MR1_mpr_an
on_fslswapdim_bet.nii.gz',
       '/home/runner/nilearn_data/oasis1/OAS1_0065_MR1/mwrc1OAS1_0065_MR1_mpr_an
on_fslswapdim_bet.nii.gz',
       '/home/runner/nilearn_data/oasis1/OAS1_0054_MR1/mwrc1OAS1_0054_MR1_mpr_an
on_fslswapdim_bet.nii.gz',
       '/home/runner/nilearn_data/oasis1/OAS1_0099_MR1/mwrc1OAS1_0099_MR1_mpr_an
on_fslswapdim_bet.nii.gz',
       '/home/runner/nilearn_data/oasis1/OAS1_0023_MR1/mwrc1OAS1_0023_MR1_mpr_an
on_fslswapdim_bet.nii.gz',
       '/home/runner/nilearn_data/oasis1/OAS1_0064_MR1/mwrc1OAS1_0064_MR1_mpr_an
on_fslswapdim_bet.nii.gz',
       '/home/runner/nilearn_data/oasis1/OAS1_0181_MR1/mwrc1OAS1_0181_MR1_mpr_an
on_fslswapdim_bet.nii.gz',
       '/home/runner/nilearn_data/oasis1/OAS1_0217_MR1/mwrc1OAS1_0217_MR1_mpr_an
on_fslswapdim_bet.nii.gz',
       '/home/runner/nilearn_data/oasis1/OAS1_0144_MR1/mwrc1OAS1_0144_MR1_mpr_an
on_fslswapdim_bet.nii.gz',
       '/home/runner/nilearn_data/oasis1/OAS1_0042_MR1/mwrc1OAS1_0042_MR1_mpr_an
on_fslswapdim_bet.nii.gz',
       '/home/runner/nilearn_data/oasis1/OAS1_0178_MR1/mwrc1OAS1_0178_MR1_mpr_an
on_fslswapdim_bet.nii.gz',
       '/home/runner/nilearn_data/oasis1/OAS1_0218_MR1/mwrc1OAS1_0218_MR1_mpr_an
on_fslswapdim_bet.nii.gz',
       '/home/runner/nilearn_data/oasis1/OAS1_0002_MR1/mwrc1OAS1_0002_MR1_mpr_an
on_fslswapdim_bet.nii.gz',
       '/home/runner/nilearn_data/oasis1/OAS1_0059_MR1/mwrc1OAS1_0059_MR1_mpr_an
on_fslswapdim_bet.nii.gz',
       '/home/runner/nilearn_data/oasis1/OAS1_0167_MR1/mwrc1OAS1_0167_MR1_mpr_an
on_fslswapdim_bet.nii.gz',
       '/home/runner/nilearn_data/oasis1/OAS1_0145_MR1/mwrc1OAS1_0145_MR1_mpr_an
on_fslswapdim_bet.nii.gz',
       '/home/runner/nilearn_data/oasis1/OAS1_0169_MR1/mwrc1OAS1_0169_MR1_mpr_an
on_fslswapdim_bet.nii.gz',
       '/home/runner/nilearn_data/oasis1/OAS1_0115_MR1/mwrc1OAS1_0115_MR1_mpr_an
on_fslswapdim_bet.nii.gz',
       '/home/runner/nilearn_data/oasis1/OAS1_0111_MR1/mwrc1OAS1_0111_MR1_mpr_an
on_fslswapdim_bet.nii.gz',
       '/home/runner/nilearn_data/oasis1/OAS1_0131_MR1/mwrc1OAS1_0131_MR1_mpr_an
on_fslswapdim_bet.nii.gz',
       '/home/runner/nilearn_data/oasis1/OAS1_0098_MR1/mwrc1OAS1_0098_MR1_mpr_an
on_fslswapdim_bet.nii.gz',
       '/home/runner/nilearn_data/oasis1/OAS1_0082_MR1/mwrc1OAS1_0082_MR1_mpr_an
on_fslswapdim_bet.nii.gz',
       '/home/runner/nilearn_data/oasis1/OAS1_0136_MR1/mwrc1OAS1_0136_MR1_mpr_an
on_fslswapdim_bet.nii.gz',
       '/home/runner/nilearn_data/oasis1/OAS1_0227_MR1/mwrc1OAS1_0227_MR1_mpr_an
on_fslswapdim_bet.nii.gz',
       '/home/runner/nilearn_data/oasis1/OAS1_0022_MR1/mwrc1OAS1_0022_MR1_mpr_an
on_fslswapdim_bet.nii.gz',
       '/home/runner/nilearn_data/oasis1/OAS1_0212_MR1/mwrc1OAS1_0212_MR1_mpr_an
on_fslswapdim_bet.nii.gz',
       '/home/runner/nilearn_data/oasis1/OAS1_0079_MR1/mwrc1OAS1_0079_MR1_mpr_an
on_fslswapdim_bet.nii.gz',
       '/home/runner/nilearn_data/oasis1/OAS1_0119_MR1/mwrc1OAS1_0119_MR1_mpr_an
on_fslswapdim_bet.nii.gz',
       '/home/runner/nilearn_data/oasis1/OAS1_0016_MR1/mwrc1OAS1_0016_MR1_mpr_an
on_fslswapdim_bet.nii.gz',
       '/home/runner/nilearn_data/oasis1/OAS1_0104_MR1/mwrc1OAS1_0104_MR1_mpr_an
on_fslswapdim_bet.nii.gz',
       '/home/runner/nilearn_data/oasis1/OAS1_0203_MR1/mwrc1OAS1_0203_MR1_mpr_an
on_fslswapdim_bet.nii.gz',
       '/home/runner/nilearn_data/oasis1/OAS1_0114_MR1/mwrc1OAS1_0114_MR1_mpr_an
on_fslswapdim_bet.nii.gz'],
      dtype='<U96')
[SpaceNetRegressor.fit] Computing mask
[SpaceNetRegressor.fit] Resampling mask
[SpaceNetRegressor.fit] Finished fit
[SpaceNetRegressor.fit] Loading data from <nibabel.nifti1.Nifti1Image object at
0x7f60cbd64220>
[SpaceNetRegressor.fit] Extracting region signals
/home/runner/work/nilearn/nilearn/.tox/doc/lib/python3.10/site-packages/nilearn/maskers/base_masker.py:137: UserWarning: Persisting input arguments took 1.29s to run.
If this happens often in your code, it can cause performance problems
(results will be correct in all cases).
The reason for this is probably some large input arguments for a wrapped
 function (e.g. large strings).
THIS IS A JOBLIB ISSUE. If you can, kindly provide the joblib's team with an
 example so that they can fix the problem.
  region_signals, aux = cache(
[SpaceNetRegressor.fit] Cleaning extracted signals
/home/runner/work/nilearn/nilearn/.tox/doc/lib/python3.10/site-packages/nilearn/maskers/base_masker.py:154: UserWarning: Persisting input arguments took 0.64s to run.
If this happens often in your code, it can cause performance problems
(results will be correct in all cases).
The reason for this is probably some large input arguments for a wrapped
 function (e.g. large strings).
THIS IS A JOBLIB ISSUE. If you can, kindly provide the joblib's team with an
 example so that they can fix the problem.
  region_signals = cache(
/home/runner/work/nilearn/nilearn/examples/02_decoding/plot_oasis_vbm_space_net.py:68: UserWarning: Brain mask (7221031.999999998 mm^3) is bigger than the standard human brain (1882989.0 mm^3).This object is probably not tuned to be used on such data.
  decoder.fit(gm_imgs_train, age_train)  # fit
[SpaceNetRegressor.fit] Mask volume = 7.22103e+06mm^3 = 7221.03cm^3
[SpaceNetRegressor.fit] Standard brain volume = 1.88299e+06mm^3
[SpaceNetRegressor.fit] Original screening-percentile: 5
[SpaceNetRegressor.fit] Corrected screening-percentile: 1.30382
[Parallel(n_jobs=2)]: Using backend LokyBackend with 2 concurrent workers.
/home/runner/work/nilearn/nilearn/.tox/doc/lib/python3.10/site-packages/joblib/externals/loky/process_executor.py:700: UserWarning: A worker stopped while some jobs were given to the executor. This can be caused by a too short worker timeout or by a memory leak.
  warnings.warn(
[Parallel(n_jobs=2)]: Done   8 out of   8 | elapsed:  1.8min finished
[SpaceNetRegressor.fit] Computing image from signals
[SpaceNetRegressor.fit] Time Elapsed: 129.96328616142273 seconds,
2.1660547693570456 minutes.
[SpaceNetRegressor.predict] Loading data from <nibabel.nifti1.Nifti1Image object
at 0x7f60cbf71b70>
[SpaceNetRegressor.predict] Extracting region signals
/home/runner/work/nilearn/nilearn/.tox/doc/lib/python3.10/site-packages/nilearn/maskers/base_masker.py:137: JobLibCollisionWarning: Cannot detect name collisions for function 'nifti_masker_extractor'
  region_signals, aux = cache(
/home/runner/work/nilearn/nilearn/.tox/doc/lib/python3.10/site-packages/nilearn/maskers/base_masker.py:137: UserWarning: Persisting input arguments took 0.86s to run.
If this happens often in your code, it can cause performance problems
(results will be correct in all cases).
The reason for this is probably some large input arguments for a wrapped
 function (e.g. large strings).
THIS IS A JOBLIB ISSUE. If you can, kindly provide the joblib's team with an
 example so that they can fix the problem.
  region_signals, aux = cache(
[SpaceNetRegressor.predict] Cleaning extracted signals
Mean square error (MSE) on the predicted age: 12.39

Visualize the decoding maps and quality of predictions

import matplotlib.pyplot as plt

from nilearn.plotting import plot_stat_map, show

# weights map
background_img = gm_imgs[0]
plot_stat_map(
    coef_img,
    background_img,
    title="graph-net weights",
    display_mode="z",
    cut_coords=1,
    cmap="cold_hot",
)

# Plot the prediction errors.
plt.figure()
plt.suptitle(f"graph-net: Mean Absolute Error {mse:.2f} years")
linewidth = 3
ax1 = plt.subplot(211)
ax1.plot(age_test, label="True age", linewidth=linewidth)
ax1.plot(y_pred, "--", c="g", label="Predicted age", linewidth=linewidth)
ax1.set_ylabel("age")
plt.legend(loc="best")
ax2 = plt.subplot(212)
ax2.plot(
    age_test - y_pred, label="True age - predicted age", linewidth=linewidth
)
ax2.set_xlabel("subject")
plt.legend(loc="best")

show()
  • plot oasis vbm space net
  • graph-net: Mean Absolute Error 12.39 years

Total running time of the script: (2 minutes 21.391 seconds)

Estimated memory usage: 3405 MB

Gallery generated by Sphinx-Gallery