def chart_prep(indexes, csv_data_list, file_name):
colors = 100 * np.random.random(100).tolist()
l = []
for my_index in indexes:
my_index_list = []
for csv_data_chunk in csv_data_list:
if csv_data_chunk[0] == my_index:
my_index_list.append(csv_data_chunk)
# l.append(my_index_list, file_name, colors.pop()))
l.append((my_index_list, my_index, colors.pop()))
return l
def plot_marker_accessions_hist(self,
pid,
marker,
m_accessions,
plot_file=None,
plot_format='png',
m_position=None,
m_chromosome=None,
title=None,
m_score=None):
"""
A histogram displaying the phenotype values (ordered) on the y-axis, and the accession on the x-axis.
"""
import matplotlib.cm as cm
import matplotlib.colors as colors
color_map = {}
colors = ['r', 'm', 'b', 'g']
proxy_rects = []
labels = []
for nt in set(marker):
c = colors.pop()
color_map[nt] = c
proxy_rects.append(plt.Rectangle((0, 0), 1, 1, fc=c, alpha=0.6))
labels.append("'%s' allele" % str(nt))
phen_values = self.get_values(pid)
ets = self.get_ecotypes(pid)
l = zip(phen_values, ets, marker)
l.sort(reverse=True)
fig = plt.figure(figsize=(20, 10))
ax = fig.add_axes([0.07, 0.15, 0.91, 0.82])
x_range = len(l) - 0.2
min_y = min(phen_values)
max_y = max(phen_values)
y_range = max_y - min_y
for i, (phen_val, accession, nt) in enumerate(l):
color = color_map[nt]
rect = ax.bar(i, phen_val, 0.8, color=color, alpha=0.6)
ax.axis([
-x_range * 0.02, x_range * 1.02, min_y - 0.05 * y_range,
max_y + 0.05 * y_range
])
ax.legend(proxy_rects, labels)
ax.set_ylabel('Phenotype value')
ax.set_xticks((sp.arange(len(ets)) + 0.4).tolist())
ax.set_xticklabels(ets, rotation='vertical', fontsize='xx-small')
ax.set_xlabel('Ecotype IDs')
fig.savefig(plot_file, format=plot_format, dpi=300)
def plot_marker_accessions_hist(
self,
pid,
marker,
m_accessions,
plot_file=None,
plot_format="png",
m_position=None,
m_chromosome=None,
title=None,
m_score=None,
):
"""
A histogram displaying the phenotype values (ordered) on the y-axis, and the accession on the x-axis.
"""
import matplotlib.cm as cm
import matplotlib.colors as colors
color_map = {}
colors = ["r", "m", "b", "g"]
proxy_rects = []
labels = []
for nt in set(marker):
c = colors.pop()
color_map[nt] = c
proxy_rects.append(plt.Rectangle((0, 0), 1, 1, fc=c, alpha=0.6))
labels.append("'%s' allele" % str(nt))
phen_values = self.get_values(pid)
ets = self.get_ecotypes(pid)
l = zip(phen_values, ets, marker)
l.sort(reverse=True)
fig = plt.figure(figsize=(20, 10))
ax = fig.add_axes([0.07, 0.15, 0.91, 0.82])
x_range = len(l) - 0.2
min_y = min(phen_values)
max_y = max(phen_values)
y_range = max_y - min_y
for i, (phen_val, accession, nt) in enumerate(l):
color = color_map[nt]
rect = ax.bar(i, phen_val, 0.8, color=color, alpha=0.6)
ax.axis([-x_range * 0.02, x_range * 1.02, min_y - 0.05 * y_range, max_y + 0.05 * y_range])
ax.legend(proxy_rects, labels)
ax.set_ylabel("Phenotype value")
ax.set_xticks((sp.arange(len(ets)) + 0.4).tolist())
ax.set_xticklabels(ets, rotation="vertical", fontsize="xx-small")
ax.set_xlabel("Ecotype IDs")
fig.savefig(plot_file, format=plot_format, dpi=300)
def draw_top(main_models,
sub_models_list,
cmap=custom_cmap,
legend_names=None,
topn=20,
offset=0.2,
width_ratio=0.1,
height_ratio=0.2,
linewidth=1,
point_size=5,
fontsize=12,
box_aspect=None,
value_aspect='auto',
id_csv_path='./data/mask/cortical_id_new.csv',
show=True,
out_path=None):
# load id Dataframe
id_df = pd.read_csv(id_csv_path, index_col=1)
main_df = models_to_dataframe(main_models)
sub_dfs = [models_to_dataframe(models) for models in sub_models_list]
sorted_main_df = main_df.sort_values('es')
top_df = sorted_main_df[:topn]
colors = cmap(np.arange(1 + len(sub_models_list))).tolist()
main_color = colors.pop(0)
fig = plt.figure(figsize=(width_ratio * topn, height_ratio * topn))
ax = fig.add_axes([0, 0, 1, 1])
legends = []
y_labels = []
y = topn
for index, row in top_df.iterrows():
ll = row['ll']
ul = row['ul']
es = row['es']
plt.scatter(es, y, s=point_size, color=main_color)
b, = plt.plot((ll, ul), (y, y), linewidth=linewidth, color=main_color)
if y == topn:
legends.append(b)
y = y - 1
y_labels.append(id_df.loc[int(index)]['name'])
i = 1
offsets = []
for i in range(2, len(sub_models_list) + 2):
if i % 2:
offsets.append(offset * int(i / 2))
else:
offsets.append(-offset * int(i / 2))
top_sub_dfs = [top_df.align(df, join='left')[1] for df in sub_dfs]
for sub_df, color, _offset in zip(top_sub_dfs, colors, offsets):
y = topn
for index, row in sub_df.iterrows():
ll = row['ll']
ul = row['ul']
es = row['es']
plt.scatter(es, y + _offset, s=point_size, color=color)
b, = plt.plot((ll, ul), (y + _offset, y + _offset),
linewidth=linewidth,
color=color)
if y == topn:
legends.append(b)
y = y - 1
i += 1
ax.set_yticks(np.arange(topn + 1))
y_labels.append('')
y_labels.reverse()
ax.set_yticklabels(y_labels, fontdict={'fontsize': fontsize})
ax.tick_params(axis='both', which='major', labelsize=fontsize)
ax.tick_params(axis='y', length=0)
if legend_names is not None and len(legend_names) == len(legends):
plt.legend(legends, legend_names)
ax.set_box_aspect(aspect=box_aspect)
ax.set_aspect(aspect=value_aspect)
if out_path is not None:
plt.savefig(out_path)
if show:
plt.show()
plt.close()
m.drawparallels(parallels,labels=[1,0,0,0],fontsize=10)
meridians = np.arange(180.,360.,10.)
m.drawmeridians(meridians,labels=[0,0,0,1],fontsize=10)
# display blue marble image (from NASA) as map background
#m.bluemarble()
conn = pypyodbc.connect('DSN=localMSSQL')
cur = conn.cursor()
sql= "select top 10000 constituentlookupid,post_code,latitude, longitude,gender from [saeed_cons_spatial] where post_code is not null"
cur.execute(sql)
dd=cur.fetchall();
df=pd.DataFrame(dd)
colors = list("rgb")
groups_gend = df.groupby(df[4])
for name, group in groups_gend:
lons=group[3].values
lats=group[2].values
gend=group[4].values
x, y = m(lons,lats)
m.scatter(x.tolist(), y.tolist(), color=colors.pop(), alpha=0.05,linewidth='0', label=name)
plt.legend()
plt.show()
