Removed sub-example due to unfit for lasso dataset - unstable float v…

…alues (#2177)

* Removed sub-example due to unfit for lasso dataset - unstable float values

* Removed remaining mention of Seitzman atlas from example

* Updated Whats whats_new

doc/whats_new.rst

@@ -67,6 +67,10 @@ Changes
 - :func:`nilearn.plotting.view_connectome` now converts NaNs in the adjacency
   matrix to 0.
 
+- Removed the plotting connectomes example which used the Seitzman atlas
+  from `examples/03_connectivity/plot_sphere_based_connectome.py`.
+  The atlas data is unsuitable for the method & the example is redundant.
+
 Fixes
 -----
 

examples/03_connectivity/plot_sphere_based_connectome.py

@@ -4,8 +4,8 @@
 
 This example shows how to extract signals from spherical regions.
 We show how to build spheres around user-defined coordinates, as well as
-centered on coordinates from the Power-264 atlas [1], the Dosenbach-160
-atlas [2], and the Seitzman-300 atlas [3].
+centered on coordinates from the Power-264 atlas [1], and the Dosenbach-160
+atlas [2].
 
 **References**
 
@@ -15,11 +15,6 @@
 [2] Dosenbach N.U., Nardos B., et al. "Prediction of individual brain maturity
 using fMRI.", 2010, Science 329, 1358-1361.
 
-[3] `Seitzman, B. A., et al. "A set of functionally-defined brain regions with
-improved representation of the subcortex and cerebellum.", 2018, bioRxiv,
-450452
-<http://doi.org/10.1101/450452>`_
-
 We estimate connectomes using two different methods: **sparse inverse
 covariance** and **partial_correlation**, to recover the functional brain
 **networks structure**.
@@ -332,55 +327,6 @@
 
 plotting.show()
 
-
-###############################################################################
-# Connectome extracted from Seitzman's atlas
-# -----------------------------------------------------
-# We repeat the same steps for Seitzman's atlas.
-seitzman = datasets.fetch_coords_seitzman_2018()
-
-coords = np.vstack((
-    seitzman.rois['x'],
-    seitzman.rois['y'],
-    seitzman.rois['z'],
-)).T
-
-###############################################################################
-# Before calculating the connectivity matrix, let's look at the distribution
-# of the regions of the default mode network.
-dmn_rois = seitzman.networks == "DefaultMode"
-dmn_coords = coords[dmn_rois]
-zero_matrix = np.zeros((len(dmn_coords), len(dmn_coords)))
-plotting.plot_connectome(zero_matrix, dmn_coords,
-                         title='Seitzman default mode network',
-                         node_color='darkred', node_size=20)
-
-###############################################################################
-# Now let's calculate connectivity for the Seitzman atlas.
-spheres_masker = input_data.NiftiSpheresMasker(
-    seeds=coords, smoothing_fwhm=6, radius=4,
-    detrend=True, standardize=True, low_pass=0.1, high_pass=0.01, t_r=2,
-    allow_overlap=True)
-
-timeseries = spheres_masker.fit_transform(func_filename,
-                                          confounds=confounds_filename)
-
-covariance_estimator = GraphicalLassoCV()
-covariance_estimator.fit(timeseries)
-matrix = covariance_estimator.covariance_
-
-plotting.plot_matrix(matrix, vmin=-1., vmax=1., colorbar=True,
-                     title='Seitzman correlation matrix')
-
-plotting.plot_connectome(matrix, coords, title='Seitzman correlation graph',
-                         edge_threshold="99.7%", node_size=20, colorbar=True)
-
-
-###############################################################################
-# We can easily identify the networks from the matrix blocks.
-print('Seitzman networks names are {0}'.format(np.unique(seitzman.networks)))
-plotting.show()
-
 ###############################################################################
 # .. seealso::
 #