49257721ec65c78965df63152b8933e9baebd4a6,examples/decoding/plot_poldrack_space_net.py,,,#,11

Before Change

////// Fit and predict ////////////////////////////////////////////////////////////////////////////////////////////////////////////////////
from nilearn.decoding import SpaceNetRegressor
penalties = ["smooth-lasso", "tv-l1"]
decoders = {}
for penalty in penalties:
    decoder = SpaceNetRegressor(mask=mask_img, penalty=penalty,
                                eps=1e-1,  // prefer large alphas
                                memory="cache", verbose=2)
    decoder.fit(X, y)  // fit
    decoders[penalty] = decoder


////// Visualization //////////////////////////////////////////////////////////////////////////////////////////////////////////////////////////
import matplotlib.pyplot as plt
from nilearn.plotting import plot_stat_map
from nilearn.image import mean_img
for penalty, decoder in decoders.iteritems():
    plot_stat_map(mean_img(decoder.coef_img_), title=penalty,
                  display_mode="yz", cut_coords=[20, -2])
plt.show()

After Change

decoder.fit(X, y)  // fit

////// Visualize TV-L1 weights
plot_stat_map(mean_img(decoder.coef_img_), title=penalty, display_mode="yz",
              cut_coords=[20, -2])

plt.show()

Instances