def cmd():
rbh_alignments = pd.read_csv(self.train_rbh_fn)
rbh_alignments['q_name'] = rbh_alignments['q_name'].apply(
fix_transdecoder_name)
print(rbh_alignments.q_name[0])
train_alignments = pd.concat([
group
for group in parsers.maf_to_df_iter(self.transcriptome_x_db_fn)
])
print(train_alignments.q_name[0])
train_alignments = train_alignments.ix[rbh_alignments.index]
self.fit(train_alignments)
data = self.preprocess(train_alignments)
data = self.transform(data)
xx, yy, Z = self.plot_grid(data.min(), data.max())
with FigureManager(filename=self.model_fn +
'.training.pdf') as (fig, ax):
ax.set_title('Boundaries and Training Data')
ax.contourf(xx,
yy,
Z,
levels=np.linspace(Z.min(), 0, 7),
cmap=plt.cm.Blues_r)
a = plt.contour(xx,
yy,
Z,
levels=[0],
linewidths=2,
colors='red')
ax.contourf(xx, yy, Z, levels=[0, Z.max()], colors='orange')
ax.scatter(data[:, 0], data[:, 1], c='white')
ax.set_xlim((data.min(), data.max()))
ax.set_ylim((data.min(), data.max()))
def test_subplots_ncols(self):
'''Test that FigureManager generates multiple subplots with ncols.
'''
with FigureManager(ncols=3) as (fig, ax):
self.assertEqual(len(ax), 3)
self.assertIsInstance(fig, figure.Figure)
for axis in ax:
self.assertIsInstance(axis, axes.Axes)
def test_save(self):
'''Test that FigureManager saves a plot.
'''
with TemporaryDirectory() as td:
with TemporaryFile(td) as filename:
with FigureManager(filename=filename + '.pdf') as (fig, ax):
ax.plot(range(10), range(10))
self.assertTrue(os.path.isfile(filename + '.pdf'))
def test_exception(self):
'''Test that FigureManager handles internal exceptions properly.
'''
with TemporaryDirectory() as td:
with TemporaryFile(td) as filename:
# now we test the manager
try:
rregexp = self.assertRaisesRegex
except AttributeError:
rregexp = self.assertRaisesRegexp
with rregexp(RuntimeError, 'TEST'):
with FigureManager(filename=filename) as (fig, ax):
raise RuntimeError('TEST')
self.assertFalse(os.path.isfile(filename),
msg='Should not have saved a file.')
def plot_crbh_fit(model_df,
hits_df,
model_plot_fn,
show=False,
figsize=(10, 10),
feature_col='E',
length_col='s_aln_len',
**fig_kwds):
plt.style.use('seaborn-ticks')
with FigureManager(model_plot_fn, show=show, figsize=figsize,
**fig_kwds) as (fig, ax):
if not len(hits_df):
return
sample_size = min(len(hits_df), 5000)
hits_df, scaled_col = scale_evalues(hits_df,
name=feature_col,
inplace=False)
ax.scatter(hits_df[length_col],
hits_df[scaled_col],
s=10,
alpha=0.7,
c=sns.xkcd_rgb['ruby'],
marker='o',
label='Query Hits')
ax.scatter(model_df['center'],
model_df['fit'],
label='CRBL Fit',
c=sns.xkcd_rgb['twilight blue'],
marker='o',
s=5,
alpha=0.7)
leg = ax.legend(fontsize='medium', scatterpoints=3, frameon=True)
leg.get_frame().set_linewidth(1.0)
ax.set_xlim(model_df['center'].min(), model_df['center'].max())
ax.set_ylim(0,
max(model_df['fit'].max(), hits_df[scaled_col].max()) + 50)
ax.set_ylabel('Score ($E_{scaled}$)' if scaled_col == 'E_scaled'\
else 'Score ({0})'.format(scaled_col))
ax.set_xlabel('Alignment Length')
def cmd():
results = []
alns = []
N = 0
y = 0
for group in parsers.maf_to_df_iter(self.transcriptome_x_db_fn):
N += len(group)
mask = self.predict(group)
results.append(group.ix[mask])
alns.append(group[self.features])
y += mask.sum()
results = pd.concat(results)
results.to_csv(self.crbl_fn)
aln_df = pd.concat(alns)
print(N, y)
data = self.preprocess(aln_df)
data = self.transform(data)
xx, yy, Z = self.plot_grid(data.min(), data.max())
with FigureManager(filename=self.model_fn + '.fitted.pdf') as (fig,
ax):
ax.set_title('Boundary and Predict Data')
ax.contourf(xx,
yy,
Z,
levels=np.linspace(Z.min(), 0, 7),
cmap=plt.cm.Blues_r)
a = plt.contour(xx,
yy,
Z,
levels=[0],
linewidths=2,
colors='red')
ax.contourf(xx, yy, Z, levels=[0, Z.max()], colors='orange')
ax.scatter(data[:, 0], data[:, 1], c='white')
ax.set_xlim((data.min(), data.max()))
ax.set_ylim((data.min(), data.max()))