Visualizing a single network¶

Open interactive notebook in Binder

Goal¶

Here we are plot the nodes on the brain from the Schaefer atlas and plot the connectivity matrix next to it

So this code may seem long but most of it is creating the data.

In [1]:

# Import packages
import templateflow.api as tf
import netplotbrain
import pandas as pd
import numpy as np
import itertools

#  
# STEP 1 - CREATE DATA
#

# +
# Get template information
template = 'MNI152NLin2009cAsym'
atlas = 'Schaefer2018'
atlas_desc = '100Parcels7Networks'

# Import packages import templateflow.api as tf import netplotbrain import pandas as pd import numpy as np import itertools # # STEP 1 - CREATE DATA # # + # Get template information template = 'MNI152NLin2009cAsym' atlas = 'Schaefer2018' atlas_desc = '100Parcels7Networks'

In [2]:

# Get and load the tsv file (information about the atlas)
atlas_path = tf.get(atlas=atlas,
                    template=template,
                    desc=atlas_desc,
                    extension='.tsv')
atlas_df = pd.read_csv(atlas_path, sep='\t')
atlas_df.head()

# Get and load the tsv file (information about the atlas) atlas_path = tf.get(atlas=atlas, template=template, desc=atlas_desc, extension='.tsv') atlas_df = pd.read_csv(atlas_path, sep='\t') atlas_df.head()

Out[2]:

	index	name	color
0	1	7Networks_LH_Vis_1	#781283
1	2	7Networks_LH_Vis_2	#781284
2	3	7Networks_LH_Vis_3	#781285
3	4	7Networks_LH_Vis_4	#781287
4	5	7Networks_LH_Vis_5	#781288

In [3]:

# Create a column called "network" by extracting the relevant information from the column "name"
atlas_df['network'] = list(map(lambda x: x.split('_')[2], atlas_df.name.values))
atlas_df.head()

# Create a column called "network" by extracting the relevant information from the column "name" atlas_df['network'] = list(map(lambda x: x.split('_')[2], atlas_df.name.values)) atlas_df.head()

Out[3]:

	index	name	color	network
0	1	7Networks_LH_Vis_1	#781283	Vis
1	2	7Networks_LH_Vis_2	#781284	Vis
2	3	7Networks_LH_Vis_3	#781285	Vis
3	4	7Networks_LH_Vis_4	#781287	Vis
4	5	7Networks_LH_Vis_5	#781288	Vis

In [4]:

# We need to create edges for this network. Let's make a 100x100 network where within network connections are 1, everything else is 0.
edges = np.zeros([100, 100])
for network in atlas_df['network'].unique():
    idx =  atlas_df[atlas_df['network']==network].index
    idx_pairs = np.array(list(itertools.combinations(idx, 2)))
    edges[idx_pairs[:, 0], idx_pairs[:, 1]] = 1
    edges[idx_pairs[:, 1], idx_pairs[:, 0]] = 1

#  
# STEP 2 - PLOT DATA
#

# We need to create edges for this network. Let's make a 100x100 network where within network connections are 1, everything else is 0. edges = np.zeros([100, 100]) for network in atlas_df['network'].unique(): idx = atlas_df[atlas_df['network']==network].index idx_pairs = np.array(list(itertools.combinations(idx, 2))) edges[idx_pairs[:, 0], idx_pairs[:, 1]] = 1 edges[idx_pairs[:, 1], idx_pairs[:, 0]] = 1 # # STEP 2 - PLOT DATA #

In [5]:

# Nodes in connectivity will automatically be organized by node_color, unless cm_order is specified.
# We set edge_alpha to 0 so they are not visible in the brain images
# If nodecolor is specified the borders of the connectivity matrix will be drawn grouping and circling them in. 
netplotbrain.plot(nodes={'template': 'MNI152NLin2009cAsym',
                         'atlas': atlas,
                         'desc': atlas_desc,
                         'resolution': 1},
                  edges=edges,
                  edge_alpha = 0,
                  view='Sc',
                  node_type='parcels',
                  nodes_df=atlas_df,
                  title=None,
                  node_color='network',
                  node_cmap='Set2')

# Nodes in connectivity will automatically be organized by node_color, unless cm_order is specified. # We set edge_alpha to 0 so they are not visible in the brain images # If nodecolor is specified the borders of the connectivity matrix will be drawn grouping and circling them in. netplotbrain.plot(nodes={'template': 'MNI152NLin2009cAsym', 'atlas': atlas, 'desc': atlas_desc, 'resolution': 1}, edges=edges, edge_alpha = 0, view='Sc', node_type='parcels', nodes_df=atlas_df, title=None, node_color='network', node_cmap='Set2')

Out[5]:

(<Figure size 432x252 with 3 Axes>, [<Axes3D: >, <Axes: >, <Axes: >])