def plot_slopes(df, fits):
fig, ax = plt.subplots()
ax.set_ylabel('slope [RFU/min]')
x = 0
x_ticks = []
x_labels = []
groups = df.groupby(['control', 'rnase_U_mL', 'well'])
for color, ((control, _, well), g) in iter_colors(sorted(groups)):
m, b, r, p, err = fits[well]
rnase_U_mL = one(uniq(g['rnase_U_mL']))
title = one(uniq(g['title']))
label = f'{title}' if control else f'{rnase_U_mL} U/mL'
if title == 'FAM':
continue
ax.plot(
[x,x],
[0,m],
color=color,
linewidth=6,
)
#ax.plot(
# [x,x],
# [0,m],
# marker='_',
# color='black',
#)
ax.errorbar(
[x],
[m],
yerr=err,
color=color,
capsize=3,
)
x_ticks.append(x)
x_labels.append(label)
x += 1
ax.set_xticks(x_ticks)
ax.set_xticklabels(x_labels, rotation='vertical')
ax.set_xlim(min(x_ticks) - 0.5, max(x_ticks) + 0.5)
ax.set_ylim(0, ax.get_ylim()[1])
fig.tight_layout()
return fig, ax
def _key_from_group(self, g, cols):
try:
key = tuple(
one(uniq(g[col]))
for col in cols
)
return self._normalize_key(key)
except ValueError:
raise KeyError(cols)
def plot_linear_fits(df, fits):
fig, axes = plt.subplots(1, 2, sharex=True, sharey=True)
axes[0].set_title('Controls')
axes[1].set_title('Experiments')
for ax in axes[:0]:
ax.set_ylabel('RFU')
for ax in axes[-1:]:
ax.set_xlabel('time [min]')
for control, gi in df.groupby(['control']):
ax = axes[0 if control else 1]
for color, (well, gj) in iter_colors(gi.groupby(['well'])):
rnase_U_mL = one(uniq(gj['rnase_U_mL']))
title = one(uniq(gj['title']))
label = f'{title}' if control else f'{rnase_U_mL} U/mL'
plot_linear_fit(ax, gj, fits[well], color=color, label=label)
ax.legend(loc='best')
return fig, axes
def format_label(self, g):
params = {
col: ','.join(map(str, uniq(g[col])))
for col in g
}
if not isinstance(self.label, list):
labels = [self.label]
else:
labels = self.labels
for label in labels:
try:
return label.format(**params)
except KeyError:
continue
def preclean(x):
x = remove_non_ascii(' '.join(i for i in uniq(x.split())))
return to_single_space(x).strip()
def preclean(x):
return strip(remove_non_ascii(
to_single_space(
' '.join(i for i in uniq(x.split()))
)))