def mds_author_term(fname1='corr_2d_mds_authors_by_terms.png', fname2='corr_2d_mds_terms_by_authors.png'):
bib_data = get_bib_data()
mat, authors, term_list, authors_cnt = get_author_by_term_mat(bib_data, tfreq=5, afreq=10)
adist = dist.squareform(dist.pdist(mat, 'correlation'))
coords,_ = mds(adist, dim=2)
fig = plt.figure()
fig.clf()
plt.xlim(-15, 20)
plt.ylim(-15, 20)
for label, x, y in zip(authors, coords[:,0], coords[:,1]):
plt.annotate(label, xy=(x*20,y*20))
plt.axis('off')
plt.savefig(fname1)
mat = mat.T
tdist = dist.squareform(dist.pdist(mat, 'correlation'))
coords, _ = mds(tdist, dim=2)
#fig = plt.figure()
fig.clf();
plt.xlim(-80,100)
plt.ylim(-100,100)
for label, x, y in zip(term_list, coords[:,0], coords[:,1]):
plt.annotate(label, xy=(x*500,y*500))
plt.axis('off')
plt.savefig(fname2)
def mds_year_term(fname1='corr_2d_mds_years_by_terms.png', fname2='corr_2d_mds_terms_by_years.png'):
bib_data = get_bib_data()
mat, years, term_list, years_cnt = get_year_by_term_mat(bib_data, freq=5)
ydist = dist.squareform(dist.pdist(mat, 'correlation'))
coords,_ = mds(ydist, dim=2)
mat = mat.T
tdist = dist.squareform(dist.pdist(mat, 'correlation'))
coords, _ = mds(tdist, dim=2)
fig = plt.figure()
fig.clf();
plt.xlim(-80,100)
plt.ylim(-100,100)
for label, x, y in zip(term_list, coords[:,0], coords[:,1]):
plt.annotate(label, xy=(x*100,y*100))
plt.axis('off')
plt.savefig(fname2)
def MDS_A_by_B(mat, labels, fname):
adist = dist.squareform(dist.pdist(mat, 'correlation'))
coords,_ = mds(adist, dim=2)
fig = plt.figure()
fig.clf()
plt.xlim(-15, 20)
plt.ylim(-15, 20)
for label, x, y in zip(labels, coords[:,0], coords[:,1]):
plt.annotate(label, xy=(x*20,y*20))
#plt.axis('off')
plt.savefig(fname)
