def plot_data_mlp_to_html(data_points, labels=[]):
# print(len(dataPoints))
data_points = cull_to_number(data_points, 250)
# print(len(dataPoints))
# print(dataPoints)
swap = swapaxes(data_points, 0, 1)
# print(swap)
timescale = []
for i, date in enumerate(swap[0]):
parsedDate = datetime.datetime.strptime(date, '%m/%d/%y %H:%M:%S')
timescale.append(parsedDate)
# timescale = [datetime.datetime.strptime(date, '%m/%d/%y %H:%M:%S') for date in swap[0]]
dates = matplotlib.dates.date2num(timescale)
fig, ax = plt.subplots(figsize=[9.28, 4.8])
plt.rcParams.update({'figure.autolayout': True})
ax.yaxis.set_major_locator(matplotlib.ticker.MultipleLocator(10))
ax.grid(which='major', axis='both', linestyle="-", alpha=0.25, linewidth=1)
ax.tick_params(labelsize='medium', width=1)
# Read labels from list
p1 = ax.plot_date(dates,
swap[1],
color='red',
linestyle='solid',
marker=None,
label=labels[0])
p2 = ax.plot_date(dates,
swap[2],
color='green',
linestyle='solid',
marker=None,
label=labels[1])
p3 = ax.plot_date(dates,
swap[3],
color='blue',
linestyle='solid',
marker=None,
label=labels[2])
s1 = ax.scatter(dates, swap[1], color='#00000000', s=50)
s2 = ax.scatter(dates, swap[2], color='#00000000', s=50)
s3 = ax.scatter(dates, swap[3], color='#00000000', s=50)
ax.legend()
# xlabels = ax.get_xticklabels()
# plt.setp(xlabels, rotation=30, horizontalalignment='right')
ax.set(ylim=[0, 100],
xlabel='Dată',
ylabel='Procent (%)',
title='Grafic umiditate plante')
# Create HTML
plugins.clear(fig)
tooltip_plugin1 = plugins.PointLabelTooltip(
s1, extract_point_position(swap[0], swap[1]))
tooltip_plugin2 = plugins.PointLabelTooltip(
s2, extract_point_position(swap[0], swap[2]))
tooltip_plugin3 = plugins.PointLabelTooltip(
s3, extract_point_position(swap[0], swap[3]))
plugins.connect(fig, tooltip_plugin1, tooltip_plugin2, tooltip_plugin3,
plugins.Zoom(), plugins.Reset())
html_str = fig_to_html(fig)
return html_str
def mpld3_enable_notebook():
"""Change the default plugins, enable ipython notebook mode and return mpld3 module."""
import mpld3
from mpld3 import plugins as plugs
plugs.DEFAULT_PLUGINS = [plugs.Reset(), plugs.Zoom(), plugs.BoxZoom(), plugs.MousePosition()]
mpld3.enable_notebook()
return mpld3
def create_plot():
fig, ax = plt.subplots()
points = ax.scatter(random_x, random_y, s=500, alpha=0.3)
plugins.clear(fig)
ax.set_title('Click info', size=14)
plugins.connect(fig, plugins.Reset(), plugins.Zoom(), ClickInfo(points))
return fig
def create_plot():
fig, ax = plt.subplots()
points = ax.scatter(np.random.rand(50), np.random.rand(50),
s=500, alpha=0.3)
plugins.clear(fig)
plugins.connect(fig, plugins.Reset(), plugins.Zoom(), ClickInfo(points))
return fig
def plot_vdot_race(data):
#CSS element for the tooltip label
css = """
div
{
font-family: Avenir, Helvetica, sans-serif;
border: 1px solid black;
padding-left: 5px;
padding-right: 5px;
text-align: center;
color: #000000;
background-color: #ffffff;
}
"""
#plt.style.use('seaborn-poster') #sets the size of the charts
plt.style.use('ggplot')
x_axis = []
y_axis = []
label = []
#print (data)
for key in data:
label.append("Race: " + str(data[key][1]) + "<br/>" + "VDOT: " +
str(data[key][3]))
y_axis.append(data[key][3])
datetime_obj = datetime.strptime(data[key][4],
'%Y-%m-%d %H:%M:%S+00:00')
x_axis.append(datetime_obj)
dates = matplotlib.dates.date2num(x_axis)
fig, ax = plt.subplots()
plt.xlabel("Race dates", fontsize=20)
plt.ylabel("VDOT Scores", fontsize=20)
plt.title("VDOT Score Over Time", fontsize=25)
line = plt.plot_date(dates,
y_axis,
linestyle='solid',
marker='.',
markersize=14)
plugins.clear(fig) # clear all plugins from the figure
tooltip = plugins.PointHTMLTooltip(line[0],
label,
hoffset=-60,
voffset=-70,
css=css)
plugins.connect(fig, plugins.Reset(), plugins.BoxZoom(), plugins.Zoom(),
tooltip)
return mpld3.fig_to_html(fig)
def make_html_plot_for_file(filename, my_title):
my_fontsize = 20
# read in csv file
if os.stat(filename).st_size:
data = np.genfromtxt(filename, delimiter=',')
else:
# empty data file
data = np.zeros([2, 2])
if len(data.shape) == 1:
return "<h2> " + my_title + " has no data </h2>"
x = data[:, 0]
fig, ax = plt.subplots(figsize=(10, 5))
for k in range(1, data.shape[1]):
y = data[:, k]
# sort data
x_sorted = [x1 for (x1, y1) in sorted(zip(x, y))]
y_sorted = [y1 for (x1, y1) in sorted(zip(x, y))]
lines = ax.plot(x_sorted,
y_sorted,
'.-',
markersize=15,
label='ion ' + str(k - 1))
plt.legend(fontsize=my_fontsize, loc='best')
plt.title(my_title, fontsize=my_fontsize)
ax.grid()
plt.xticks(fontsize=my_fontsize)
plt.yticks(fontsize=my_fontsize)
plt.xlim([np.min(x), np.max(x)])
#90% of the time that this function takes is taken up after this line
plugins.clear(fig)
plugins.connect(fig, plugins.Reset(), plugins.BoxZoom(),
plugins.Zoom(enabled=True),
plugins.MousePosition(fontsize=my_fontsize))
java_txt = mpld3.fig_to_html(fig)
plt.close()
return java_txt
def spladder_viz(options):
"""Main visualization code"""
### parse parameters from options object
options = settings.parse_args(options, identity='viz')
### create plot directory if it does not exist yet
dirname = options.outdir if options.testdir == '-' else options.testdir
if not os.path.exists(os.path.join(dirname, 'plots')):
os.mkdir(os.path.join(dirname, 'plots'))
if options.format == 'd3':
try:
import mpld3
from mpld3 import plugins
except ImportError:
sys.stderr.write(
"ERROR: missing package for output format d3. Package mpld3 required"
)
sys.exit(1)
### load gene information
all_gene_names = get_gene_names(options.outdir, options.confidence,
options.validate_sg, options.verbose)
### set color maps
cmap_cov = plt.get_cmap('jet')
cmap_edg = plt.get_cmap('jet')
### Data Range
RangeData = namedtuple('RangeData', ['chr', 'start', 'stop'])
plots = []
test_tag = []
### if the --test option was given, populate the options object with additional track fields
### otherwise just create a single plot
if options.test is None:
plots.append(options.data_tracks)
test_tag.append('')
else:
for test_info in options.test:
_parse_test_info(test_info, options.outdir, options.confidence,
plots, test_tag)
for plot_data in plots:
if len(options.data_tracks) > 0:
plot_data.extend(options.data_tracks)
### generate all plots to be completed
for p, plot_data in enumerate(plots):
### get range information
genes, gene_names, gids = [], [], []
events, eids = [], []
coords = []
for range_info in options.range:
### genes
if range_info[0] == 'gene':
_parse_gene_info(range_info[1:], genes, gene_names, gids,
all_gene_names, options.outdir,
options.confidence, options.validate_sg,
options.verbose)
### events
elif range_info[0] == 'event':
_parse_event_info(eids, gids, range_info[1:], events,
options.outdir, options.confidence,
options.verbose)
### coordinate ranges
elif range_info[0] == 'coordinate':
coords.append(
RangeData._make([
range_info[1],
int(range_info[2]),
int(range_info[3])
]))
### check data tracks for additional range information if no --range was given
if len(genes) + len(events) + len(coords) == 0:
for range_info in plot_data:
### splicegraph / transcripts
if range_info[0] in ['splicegraph', 'transcript']:
_parse_gene_info(range_info[1:], genes, gene_names, gids,
all_gene_names, options.outdir,
options.confidence, options.validate_sg,
options.verbose)
### events
elif range_info[0] == 'event':
_parse_event_info(eids, gids, range_info[1:], events,
options.outdir, options.confidence,
options.verbose)
### gene (this is only needed internally for --test mode)
elif range_info[0] == 'gene':
_parse_gene_info(range_info[1:], genes, gene_names, gids,
all_gene_names, options.outdir,
options.confidence, options.validate_sg,
options.verbose)
### check that everything is on the same chromosome
plotchrm = np.unique([_.chr for _ in events + genes + coords])
if plotchrm.shape[0] > 1:
sys.stderr.write(
'ERROR: the provided gene/event/coordinate ranges are on different chromosomes and canot be plotted jointly\n'
)
sys.exit(1)
### identify the plotting range
plotrange, plotrange_orig = _set_plotting_range(genes, events, coords)
### parse all elements to be plotted as data tracks
data_tracks = []
for data_element in plot_data:
track_types = data_element[0].split(',')
for track_type in track_types:
data_tracks.append(dt.DataTrack(track_type, data_element[1:]))
### set up figure and layout
gs = gridspec.GridSpec(len(data_tracks), 1, hspace=0.05)
fig = plt.figure(figsize=(15, 3 * len(data_tracks)), dpi=200)
axes = []
for i, data_track in enumerate(data_tracks):
### plot splicing graph
if data_track.type == 'splicegraph':
ax = _add_ax(axes, fig, gs)
if len(data_track.event_info) == 0:
_genes = genes
_gene_names = gene_names
else:
_genes, _gene_names, _gids = [], [], []
_parse_gene_info(range_info[1:], _genes, _gene_names,
_gids, all_gene_names, options.outdir,
options.confidence, options.validate_sg,
options.verbose)
for gene in _genes:
gene.from_sparse()
plot_graph(gene.splicegraph.vertices,
gene.splicegraph.edges,
ax,
xlim=plotrange,
label=gene.name)
gene.to_sparse()
ax.set_ylabel('splicing graph')
ax.get_yaxis().set_label_coords(1.02, 0.5)
### plot annotated transcripts
if data_track.type == 'transcript':
ax = _add_ax(axes, fig, gs)
if len(data_track.event_info) == 0:
_genes = genes
_gene_names = gene_names
else:
_genes, _gene_names, _gids = [], [], []
_parse_gene_info(range_info[1:], _genes, _gene_names,
_gids, all_gene_names, options.outdir,
options.confidence, options.validate_sg,
options.verbose)
for gene in _genes:
gene.from_sparse()
multiple(gene.exons,
ax=ax,
x_range=plotrange,
labels=gene.transcripts,
grid=True)
gene.to_sparse()
ax.set_ylabel('transcripts')
ax.get_yaxis().set_label_coords(1.02, 0.5)
### plot events
if data_track.type == 'event':
ax = _add_ax(axes, fig, gs)
### no event ids given - plot the one from range
if len(data_track.event_info) == 0:
_events = events
### events are given in the track - plot those instead
else:
_eids, _events = [], []
_parse_event_info(_eids, gids, data_track.event_info,
_events, options.outdir,
options.confidence, options.verbose)
event_list = [[event.exons1, event.exons2]
for event in _events]
labels = ['_'.join([_.event_type, str(_.id)]) for _ in _events]
multiple(event_list,
ax=ax,
x_range=plotrange,
color='green',
padding=None,
grid=True,
labels=labels)
vax.clean_axis(ax, allx=True)
ax.set_ylabel('events')
ax.get_yaxis().set_label_coords(1.02, 0.5)
### plot coverage tracks
if data_track.type == 'coverage':
ax = _add_ax(axes, fig, gs)
min_sample_size = 5 ### TODO make that an option
caxes = []
labels = []
norm = plt.Normalize(0, len(data_track.bam_fnames))
for g, bam_group in enumerate(data_track.bam_fnames):
label = 'group %i' % (g + 1) if len(
data_track.group_labels
) == 0 else data_track.group_labels[g]
caxes.append(
cov_from_bam(plotchrm[0],
plotrange[0],
plotrange[1],
bam_group,
subsample=min_sample_size,
ax=ax,
intron_cnt=True,
log=options.log,
xlim=plotrange,
color_cov=cmap_cov(norm(g)),
color_intron_edge=cmap_edg(norm(g)),
grid=True,
min_intron_cnt=options.mincount,
return_legend_handle=True,
label=label))
labels.append(label)
ax.get_yaxis().set_label_coords(1.02, 0.5)
if len(caxes) > 0:
ax.legend(caxes, labels)
### plot segment counts
if data_track.type == 'segments':
ax = _add_ax(axes, fig, gs)
for g, gid in enumerate(gids):
(segments, edges, edge_idx,
strains) = get_seg_counts(gid, options.outdir,
options.confidence,
options.validate_sg)
seg_sample_idx = None
if len(data_track.strains) > 0:
seg_sample_idx = []
for track_strains in data_track.strains:
seg_sample_idx.append(
np.where(np.in1d(strains, track_strains))[0])
cov_from_segments(genes[g],
segments,
edges,
edge_idx,
ax,
xlim=plotrange,
log=options.log,
grid=True,
order='C',
sample_idx=seg_sample_idx)
ax.get_yaxis().set_label_coords(1.02, 0.5)
ax.set_ylabel('segment counts')
### set x axis ticks
for i, ax in enumerate(axes):
if i == len(axes) - 1:
ax.set_xticks(
np.around(
np.linspace(plotrange_orig[0], plotrange_orig[1], 10)))
ax.set_xlabel('chromosome ' + plotchrm[0])
else:
ax.set_xticks([])
### save plot into file
if options.format == 'd3':
out_fname = os.path.join(
dirname, 'plots', '%s%s.html' % (options.outbase, test_tag[p]))
plugins.clear(fig)
plugins.connect(fig, plugins.Zoom(enabled=True))
mpld3.save_html(fig, open(out_fname, 'w'))
else:
out_fname = os.path.join(
dirname, 'plots',
'%s%s.%s' % (options.outbase, test_tag[p], options.format))
plt.savefig(out_fname, format=options.format, bbox_inches='tight')
plt.close(fig)
def spladder_viz():
"""Main visualization code"""
### parse command line parameters
options = parse_options(sys.argv)
### create plot directory if it does not exist yet
if not os.path.exists(os.path.join(options.outdir, 'plots')):
os.mkdir(os.path.join(options.outdir, 'plots'))
if options.format == 'd3':
try:
import mpld3
from mpld3 import plugins
except ImportError:
sys.stderr.write(
"ERROR: missing package for format d3. Package mpld3 required")
sys.exit(1)
### load gene information
genes = load_genes(options)
rows = get_plot_len(options)
gs = gridspec.GridSpec(rows, 1)
### get coloring
cmap_cov = plt.get_cmap('jet')
cmap_edg = plt.get_cmap('jet')
### plot log scale?
log_tag = ''
if options.log:
log_tag = '.log'
event_tag = ''
### did we get any labels?
if options.labels != '-':
options.labels = options.labels.strip(',').split(',')
assert len(options.labels) == len(
options.bams.strip(':').split(':')
), "The number of given labels (%i) needs to match the number of given bam file groups (%i)" % (
len(options.labels), len(options.bams.strip(':').split(':')))
### the user specified a gene to plot
if options.gene_name is not None:
gid = sp.where(
sp.array([x.name.split('.')[0]
for x in genes]) == options.gene_name.split('.')[0])[0]
if gid.shape[0] == 0:
sys.stderr.write(
'ERROR: provided gene ID %s could not be found, please check for correctness\n'
% options.gene_name)
sys.exit(1)
gids = [
sp.where(sp.array([x.name
for x in genes]) == options.gene_name)[0][0]
]
### no gene specified but result provided - plot all genes with confirmed events
### if an event_id is provided, only the associated gene will be plotted
else:
gids = get_gene_ids(options)
### iterate over genes to plot
for gid in gids:
if options.format == 'd3':
fig = plt.figure(figsize=(12, 2 * rows), dpi=100)
else:
fig = plt.figure(figsize=(18, 3 * rows), dpi=200)
axes = []
### gather information about the gene we plot
gene = get_gene(genes[gid])
if options.verbose:
print 'plotting information for gene %s' % gene.name
### plot splicing graph
axes.append(fig.add_subplot(gs[len(axes), 0]))
plot_graph(gene.splicegraph.vertices, gene.splicegraph.edges, axes[-1])
xlim = axes[-1].get_xlim()
axes[-1].set_title('Splicing graph for %s' % gene.name)
### plot annotated transcripts
if options.transcripts:
axes.append(fig.add_subplot(gs[len(axes), 0], sharex=axes[0]))
multiple(gene.exons, ax=axes[-1], x_range=xlim)
axes[-1].set_title('Annotated Transcripts')
### plot coverage information for a set of samples
if options.bams != '-':
samples = options.bams.strip(':').split(':')
plot_bam(options, gene, samples, fig, axes, gs, xlim, cmap_cov,
cmap_edg)
### plot all the samples in a single plot
if len(samples) > 1:
plot_bam(options,
gene,
samples,
fig,
axes,
gs,
xlim,
cmap_cov,
cmap_edg,
single=False)
### plot segment counts
(segments, edges, edge_idx) = get_seg_counts(options, gid)
axes.append(fig.add_subplot(gs[len(axes), 0], sharex=axes[0]))
if identity() == 'matlab':
cov_from_segments(gene,
segments,
edges,
edge_idx,
axes[-1],
xlim=xlim,
log=options.log,
grid=True,
order='F')
else:
cov_from_segments(gene,
segments,
edges,
edge_idx,
axes[-1],
xlim=xlim,
log=options.log,
grid=True,
order='C')
axes[-1].set_title('Segment counts')
### plot structure of a single given event
#if options.event_id is not None:
axes.append(fig.add_subplot(gs[len(axes), 0], sharex=axes[0]))
### user defined event to plot
if options.event_id is not None:
event_info = [
x[::-1] for x in re.split(
r'[._]', options.event_id[::-1], maxsplit=1)[::-1]
]
event_info[1] = int(event_info[1]) - 1
event_info = sp.array(event_info, dtype='str')[sp.newaxis, :]
event_tag = '.%s' % options.event_id
### get all significant events of the current gene
else:
event_info = get_conf_events(options, gid)
plot_event(options, event_info, axes[-1], xlim)
### we only need to adapt the xoom for one axis object - as we share the x
zoom_x = [float(x) for x in options.zoom_x.split(',')]
xlim = axes[0].get_xlim()
xdiff = xlim[1] - xlim[0]
axes[0].set_xlim(
[xlim[0] + (zoom_x[0] * xdiff), xlim[0] + (zoom_x[1] * xdiff)])
plt.tight_layout()
### save plot into file
if options.format == 'd3':
out_fname = os.path.join(
options.outdir, 'plots', 'gene_overview_C%i_%s%s%s.html' %
(options.confidence, gene.name, event_tag, log_tag))
plugins.clear(fig)
plugins.connect(fig, plugins.Zoom(enabled=True))
mpld3.save_html(fig, open(out_fname, 'w'))
else:
out_fname = os.path.join(
options.outdir, 'plots', 'gene_overview_C%i_%s%s%s.%s' %
(options.confidence, gene.name, event_tag, log_tag,
options.format))
plt.savefig(out_fname, format=options.format, bbox_inches='tight')
plt.close(fig)
def plot_beats_per_mile(data):
#divide meters by this factor to get in terms of miles
convert_meters_to_miles = 1609.344
#plt.style.use('seaborn-poster') #sets the size of the charts
plt.style.use('ggplot')
#Data is dictionary mapping id to (activity_name, start_date_local, average_hr, distance, total_elevation_gain)
data_by_week = {}
for key in data:
datetime_obj = datetime.strptime(data[key][1], '%Y-%m-%d %H:%M:%S')
year_week_val = datetime_obj.isocalendar()[:-1]
if year_week_val not in data_by_week:
#Storing average_hr, distance, total_elevation_gain
data_by_week[year_week_val] = [(data[key][2], data[key][3],
data[key][4])]
else:
data_by_week[year_week_val].append(
(data[key][2], data[key][3], data[key][4]))
x_axis = []
x_label = []
y_axis = []
for key in data_by_week:
print(key)
total_hr_each_week = 0
distance_week = 0
for activity in data_by_week[key]:
total_hr_each_week += activity[0] * (activity[1] /
convert_meters_to_miles)
distance_week += activity[1] / convert_meters_to_miles
avg_hr_per_mile = total_hr_each_week / distance_week
x_label.append(str(key[0]) + "-" + str(key[1]))
x_axis.append(key[0] * 52 + key[1])
y_axis.append(avg_hr_per_mile)
total_hr_each_week = 0
distance_week = 0
fig, ax = plt.subplots()
plt.xlabel("Weeks", fontsize=20)
plt.ylabel("Average Heartrate Per Mile", fontsize=20)
plt.title("Average Heartrate Per Mile Sorted by Week", fontsize=25)
line = plt.plot(x_axis,
y_axis,
linestyle='solid',
marker='.',
markersize=14)
x = np.asarray(x_axis)
y = np.asarray(y_axis)
b, m = polyfit(x, y, 1)
#Plots regression line
plt.plot(x, b + m * x, linestyle='solid')
plugins.clear(fig) # clear all plugins from the figure
#tooltip = plugins.PointHTMLTooltip(line[0], label,
#hoffset = -60, voffset = -70, css = css)
plugins.connect(fig, plugins.Reset(), plugins.BoxZoom(), plugins.Zoom())
return mpld3.fig_to_html(fig)
plot_rest = jitter(
reduced[rest["Membership"]]["PC1"],
reduced[rest["Membership"]]["PC2"],
c="b",
alpha=0.4,
label="Non-WMC"
)
plot_wmc = jitter(
reduced[wmc["Membership"]]["PC1"],
reduced[wmc["Membership"]]["PC2"],
c="g",
alpha=0.4,
label="WMC"
)
plt.xlabel("PC1")
plt.ylabel("PC2")
tooltip_rest = plugins.PointHTMLTooltip(plot_rest, reduced[rest["Membership"]].index.tolist())
tooltip_wmc = plugins.PointHTMLTooltip(plot_wmc, reduced[wmc["Membership"]].index.tolist())
zoom = plugins.Zoom(button=False, enabled=True)
box_zoom = plugins.BoxZoom(button=False, enabled=True)
plugins.connect(plt.gcf(), tooltip_rest)
plugins.connect(plt.gcf(), tooltip_wmc)
mpld3.save_html(plt.gcf(), "pca-2017.html")
def spladder_viz(options):
"""Main visualization code"""
### parse parameters from options object
options = settings.parse_args(options, identity='viz')
### create plot directory if it does not exist yet
if options.testdir != '-':
dirname = options.testdir
else:
dirname = options.outdir
if not os.path.exists(os.path.join(dirname, 'plots')):
os.mkdir(os.path.join(dirname, 'plots'))
if options.format == 'd3':
try:
import mpld3
from mpld3 import plugins
except ImportError:
sys.stderr.write(
"ERROR: missing package for output format d3. Package mpld3 required"
)
sys.exit(1)
### load gene information
gene_names = get_gene_names(options)
rows = get_plot_len(options)
gs = gridspec.GridSpec(rows, 1)
### set color maps
cmap_cov = plt.get_cmap('jet')
cmap_edg = plt.get_cmap('jet')
### plot log scale?
log_tag = ''
if options.log:
log_tag = '.log'
event_tag = ''
### did we get any labels?
if ',' in options.labels:
options.labels = options.labels.strip(',').split(',')
assert len(options.labels) == len(
options.bam_fnames
), "The number of given labels (%i) needs to match the number of given bam file groups (%i)" % (
len(options.labels), len(options.bam_fnames))
# Collect genes to be plotted - there are three cases possible
# 1) the user provides a gene ID
# 2) all genes containing a significant event of a given type from differential testing are plotted
# 3) all genes that contain any event are plotted
### the user chose a specific gene for plotting
### create pairs of gene ids and an event_id (the latter is None by default)
if options.gene_name is not None:
gids = [[
sp.where(sp.array(gene_names) == options.gene_name)[0][0],
options.event_id
]]
if len(gids) == 0:
sys.stderr.write(
'ERROR: provided gene ID %s could not be found, please check for correctness\n'
% options.gene_name)
sys.exit(1)
### the plotting happens on the results of spladder test
### the user chooses to plot the top k significant events
### this requires the event type to be specified
elif options.test_result > 0:
gene_names = []
for event_type in options.event_types:
### the testing script should generate a setup file for the test
### SETUP is structured as follows:
### [gene_strains, event_strains, dmatrix0, dmatrix1, event_type, options]
labels = options.test_labels.split(':')
options.labels = labels
if options.testdir != '-':
testdir = dirname
else:
testdir = os.path.join(
dirname, 'testing_%s_vs_%s' % (labels[0], labels[1]))
SETUP = pickle.load(
open(
os.path.join(
testdir, 'test_setup_C%i_%s.pickle' %
(options.confidence, event_type)), 'rb'))
### get strains to plot
idx1 = sp.where(sp.in1d(SETUP[0], SETUP[6]['conditionA']))[0]
idx2 = sp.where(sp.in1d(SETUP[0], SETUP[6]['conditionB']))[0]
### load test results
for l, line in enumerate(
open(
os.path.join(
testdir, 'test_results_C%i_%s.tsv' %
(options.confidence, event_type)), 'r')):
if l == 0:
continue
if l > options.test_result:
break
sl = line.strip().split('\t')
gene_names.append([sl[1], sl[0]])
gids = get_gene_ids(options, gene_names)
### no gene specified but result provided - plot all genes with confirmed events
### if an event_id is provided, only the associated gene will be plotted
else:
gids = get_gene_ids(options)
### iterate over genes to plot
for gid in gids:
### gather information about the gene we plot
gene = load_genes(options, idx=[gid[0]])[0]
if options.verbose:
print('plotting information for gene %s' % gene.name)
gene.from_sparse()
### event to plot is specified with the gene id list
if gid[1] is not None:
event_info = [
x[::-1]
for x in re.split(r'[._]', gid[1][::-1], maxsplit=1)[::-1]
]
event_info[1] = int(event_info[1]) - 1
event_info = sp.array(event_info, dtype='str')[sp.newaxis, :]
event_tag = '.%s' % gid[1]
### get all confident events of the current gene
else:
event_info = get_conf_events(options, gid[0])
### go over different plotting options
axes = []
### plot result of testing
if options.test_result > 0:
fig = plt.figure(figsize=(9, 5), dpi=200)
gs = gridspec.GridSpec(2, 1, height_ratios=[4, 1])
_add_ax(fig, axes, gs)
_add_ax(fig, axes, gs)
_plot_event(options,
event_info,
fig,
axes[1],
gs,
None,
padding=100)
start, stop = axes[1].get_xlim()
plot_bam(options,
gene,
options.bam_fnames,
fig,
axes[0],
gs,
None,
cmap_cov,
cmap_edg,
single=False,
sharex=axes[1],
start=int(start),
stop=int(stop))
### plot custom layout
else:
### set defaults
if not options.user:
options.splicegraph = True
options.transcripts = True
if options.format == 'd3':
fig = plt.figure(figsize=(12, 2 * rows), dpi=100)
else:
fig = plt.figure(figsize=(18, 3 * rows), dpi=200)
xlim = None
### plot splicing graph
if options.splicegraph:
_plot_splicegraph(gene, fig, axes, gs)
xlim = axes[-1].get_xlim()
### plot annotated transcripts
if options.transcripts:
sharex = None if len(axes) == 0 else axes[0]
axes.append(fig.add_subplot(gs[len(axes), 0], sharex=sharex))
multiple(gene.exons, ax=axes[-1], x_range=xlim)
axes[-1].set_title('Annotated Transcripts')
### plot coverage information for a set of given samples
if len(options.bam_fnames) > 0:
plot_bam(options, gene, options.bam_fnames, fig, axes, gs,
xlim, cmap_cov, cmap_edg)
### plot all the samples in a single plot
if len(options.bam_fnames) > 1:
plot_bam(options,
gene,
options.bam_fnames,
fig,
axes,
gs,
xlim,
cmap_cov,
cmap_edg,
single=False)
### plot segment counts
if len(options.bam_fnames
) == 0 or False: # add option for segment plots
if options.test_result > 0:
_plot_segments(options, gid, fig, axes, gs, [idx1, idx2])
else:
_plot_segments(options, gid, fig, axes, gs)
### plot structure of a single given event
_plot_event(options, event_info, fig, axes, gs, xlim)
### we only need to adapt the xoom for one axis object - as we share the x
zoom_x = [float(x) for x in options.zoom_x.split(',')]
xlim = axes[0].get_xlim()
xdiff = xlim[1] - xlim[0]
axes[0].set_xlim(
[xlim[0] + (zoom_x[0] * xdiff), xlim[0] + (zoom_x[1] * xdiff)])
for ax in axes:
vax.clean_axis(ax)
plt.tight_layout()
### save plot into file
if options.format == 'd3':
out_fname = os.path.join(
dirname, 'plots', 'gene_overview_C%i_%s%s%s.html' %
(options.confidence, gene.name, event_tag, log_tag))
plugins.clear(fig)
plugins.connect(fig, plugins.Zoom(enabled=True))
mpld3.save_html(fig, open(out_fname, 'w'))
else:
if options.test_result > 0:
out_fname = os.path.join(
dirname, 'plots', 'gene_overview_C%i_%s%s%s.%s' %
(options.confidence, gene.name, event_tag, log_tag,
options.format))
else:
out_fname = os.path.join(
dirname, 'plots', 'gene_overview_C%i_%s%s%s.%s' %
(options.confidence, gene.name, event_tag, log_tag,
options.format))
plt.savefig(out_fname, format=options.format, bbox_inches='tight')
plt.close(fig)
def result():
try:
g.conn = sqlite3.connect('case_data.db')
g.curr = g.conn.cursor()
g.curr.execute("SELECT country_name FROM interval_tb")
country = g.curr.fetchone()[0]
g.curr.execute("SELECT duration FROM interval_tb")
period = g.curr.fetchall()
g.curr.execute("SELECT restriction FROM interval_tb")
restrictions = g.curr.fetchall()
g.curr.execute("SELECT * FROM cases_tb WHERE country_name=?",
(country, ))
country_statistics = g.curr.fetchone()
g.conn.commit()
g.conn.close()
plt.rcParams.update({'font.size': 12})
def differential(initial, time, *args):
# Holds initial conditions for the population
susceptible, exposed, infected, recovered = initial
# System of differential equations
# Represents changes in the susceptible, exposed, infected, and recovered population
dSdt = -(beta * susceptible * infected) / POPULATION_SIZE
dEdt = beta * susceptible * infected / POPULATION_SIZE - SIGMA * exposed
dIdt = SIGMA * exposed - GAMMA * infected
dRdt = GAMMA * infected
return dSdt, dEdt, dIdt, dRdt
POPULATION_SIZE = country_statistics[3]
BETA = 0.956
SIGMA = 1 / 5.2
GAMMA = 1 / 2.3
REPRODUCTION_NUMBER = BETA / GAMMA
period = list(period)
time = [0]
restrictions = list(restrictions)
contact_state = []
for i in range(len(period)):
period[i] = period[i][0]
for i in range(len(restrictions)):
restrictions[i] = restrictions[i][0]
time += list(accumulate(period))
for j in range(len(period)):
if restrictions[j] == "Lockdown":
contact_state.append(0.26)
elif restrictions[j] == "Vacation":
contact_state.append(0.46)
elif restrictions[j] == "School closure":
contact_state.append(0.8)
else:
contact_state.append(1)
susceptible_initial = POPULATION_SIZE - \
country_statistics[2] - country_statistics[1]
exposed_initial = 0
infected_initial = country_statistics[2]
recovered_initial = country_statistics[1]
infected_population = 0
current_maximum = 0
fig, ax = plt.subplots()
for graph in range(len(period)):
beta = BETA * contact_state[graph]
duration = np.linspace(time[graph], time[graph + 1],
period[graph] * 2 + 1)
solution = odeint(differential,
(susceptible_initial, exposed_initial,
infected_initial, recovered_initial),
duration,
args=(beta, SIGMA, GAMMA, POPULATION_SIZE))
susceptible_initial = solution[period[graph] * 2, 0]
exposed_initial = solution[period[graph] * 2, 1]
infected_initial = solution[period[graph] * 2, 2]
recovered_initial = solution[period[graph] * 2, 3]
current_maximum = max(solution[:, 2])
if current_maximum > infected_population:
infected_population = current_maximum
if graph == 0:
ax.plot(duration, solution[:, 0], 'r', label='Susceptible')
ax.plot(duration, solution[:, 1], 'g', label='Exposed')
ax.plot(duration, solution[:, 2], 'b', label='Infectious')
ax.plot(duration, solution[:, 3], 'y', label='Recovered')
else:
ax.plot(duration, solution[:, 0], 'r')
ax.plot(duration, solution[:, 1], 'g')
ax.plot(duration, solution[:, 2], 'b')
ax.plot(duration, solution[:, 3], 'y')
ax.set_xlim(0, time[-1])
ax.set_ylim(0, POPULATION_SIZE + POPULATION_SIZE * 0.01)
ax.legend(loc='best')
ax.set_xlabel('Time (days)')
ax.set_ylabel('Population', rotation='horizontal')
fig.set_size_inches(8, 6)
ax.yaxis.set_label_coords(-0.1, 1.02)
plugins.clear(fig)
plugins.connect(fig, plugins.Reset())
plugins.connect(fig, plugins.BoxZoom(enabled=True))
plugins.connect(fig, plugins.Zoom())
seir_graph = mpld3.fig_to_html(fig)
plt.close()
effective_reproduction_number = REPRODUCTION_NUMBER * \
susceptible_initial/POPULATION_SIZE
effective_reproduction_number = float(
"{:0.3f}".format(effective_reproduction_number))
return {
'graph': seir_graph,
'r0': REPRODUCTION_NUMBER,
'rt': effective_reproduction_number,
'susceptible_state': int(susceptible_initial),
'exposed_state': int(exposed_initial),
'infected_state': int(infected_initial),
'recovered_state': int(recovered_initial),
'max_active_cases': int(infected_population)
}
except:
return jsonify(0)
def problem_view(request, problem_id):
user = request.user
problem = get_object_or_404(Problem, id=problem_id)
if problem.user_id != user and not user.is_superuser:
raise PermissionDenied
parse_error = False
a = None
b = None
c = None
a_copy = None
b_copy = None
result = None
figure = None
parse_result = parse_problem(problem.problem_text)
print parse_result
if parse_result is not None:
a, b, c, x = parse_result
if a and b and c and x:
result = simple_simplex(copy.deepcopy(a), copy.deepcopy(b),
copy.deepcopy(c))
b.pop(0)
a_copy = copy.deepcopy(a)
b_copy = copy.deepcopy(b)
if len(x) is 2:
tmp = [0, 0]
tmp[1] = 1
a.insert(0, copy.deepcopy(tmp))
tmp[0] = 1
tmp[1] = 0
a.insert(0, copy.deepcopy(tmp))
b.insert(0, 0)
b.insert(0, 0)
dots = []
for i in range(len(b)):
dots.append([])
shape_x = []
for i in xrange(0, len(b)):
for j in xrange(i + 1, len(b)):
if a[i][0] == a[j][0] and a[i][1] == a[j][1]:
continue
shape_a = np.array([[a[i][0], a[i][1]],
[a[j][0], a[j][1]]])
shape_b = np.array([b[i], b[j]])
res = np.linalg.solve(shape_a, shape_b)
dots[i].append(res)
dots[j].append(res)
# fig, ax = plt.subplots()
fig = plt.figure()
ax = fig.add_subplot(111)
for dot in dots:
ax.plot(
[float(dot[0][0]), float(dot[1][0])],
[float(dot[0][1]), float(dot[1][1])],
ls='-',
color='green',
marker='o',
lw=2)
# print([dot[0], dot[1]])
ax.set_xlabel('X axis')
ax.set_ylabel('Y axis')
ax.set_title('Plot of 2D Problem!', size=14)
plugins.clear(fig)
plugins.connect(fig, plugins.Reset(),
plugins.Zoom(enabled=True), plugins.BoxZoom())
figure = fig_to_html(fig,
d3_url=STATIC_URL + 'js/d3.min.js',
mpld3_url=STATIC_URL + 'js/mpld3.v0.2.js',
use_http=True)
else:
figure = 'Larger than 2D!'
else:
parse_error = True
return render_to_response(
'problems/view.html', {
'user': user,
'problem': problem,
'parse_error': parse_error,
'figure': figure,
'a': a_copy,
'b': b_copy,
'c': c,
'result': result,
'view_name': 'Problem - %s' % problem.id,
})
def render(self, idf, filename='output.html'):
progress_inst = helpers.Progress(idf.opt)
self.progress = progress_inst.progress
if idf.opt['outputFilename']:
filename = idf.opt['outputFilename']
if idf.opt['outputAs'] == 'html':
# write matplotlib/d3 plots to html file
import matplotlib
import matplotlib.pyplot as plt, mpld3
import matplotlib.axes
from mpld3 import plugins
from jinja2 import Environment, FileSystemLoader
elif idf.opt['outputAs'] in ['pdf', 'interactive', 'tikz']:
# show plots in separate matplotlib windows
import matplotlib
if idf.opt['outputAs'] == 'pdf':
from matplotlib.backends.backend_pdf import PdfPages
pp = PdfPages(filename)
import matplotlib.pyplot as plt
import matplotlib.axes
else:
print("No proper output method given. Not plotting.")
return
font_size = 10
if idf.opt['outputAs'] in ['pdf', 'tikz']:
if idf.opt['plotPerJoint']:
font_size = 30
else:
font_size = 12
matplotlib.rcParams.update({'font.size': font_size})
matplotlib.rcParams.update({'axes.labelsize': font_size - 5})
matplotlib.rcParams.update({'axes.linewidth': font_size / 15.})
matplotlib.rcParams.update({'axes.titlesize': font_size - 2})
matplotlib.rcParams.update({'legend.fontsize': font_size - 2})
matplotlib.rcParams.update({'xtick.labelsize': font_size - 5})
matplotlib.rcParams.update({'ytick.labelsize': font_size - 5})
matplotlib.rcParams.update({'lines.linewidth': font_size / 15.})
matplotlib.rcParams.update({'patch.linewidth': font_size / 15.})
matplotlib.rcParams.update({'grid.linewidth': font_size / 20.})
# skip some samples so graphs don't get too large/detailed TODO: change skip so that some
# maximum number of points is plotted (determined by screen etc.)
skip = 5
#create figures and plots
figures = list()
for ds in self.progress(range(len(self.datasets))):
group = self.datasets[ds]
fig, axes = plt.subplots(len(group['dataset']),
sharex=True,
sharey=True)
# scale unified scaling figures to same ranges and add some margin
if group['unified_scaling']:
ymin = 0
ymax = 0
for i in range(len(group['dataset'])):
ymin = np.min(
(np.min(group['dataset'][i]['data']), ymin)) * 1.05
ymax = np.max(
(np.max(group['dataset'][i]['data']), ymax)) * 1.05
#plot each group of data
for d_i in range(len(group['dataset'])):
d = group['dataset'][d_i]
if not issubclass(type(axes), matplotlib.axes.SubplotBase):
ax = axes[d_i]
else:
ax = axes
axes = [axes]
if idf.opt['outputAs'] != 'tikz':
ax.set_title(d['title'])
if group['unified_scaling']:
ax.set_ylim([ymin, ymax])
for data_i in range(0, len(d['data'])):
if len(d['data'][data_i].shape) > 1:
#data matrices
for i in range(0, d['data'][data_i].shape[1]):
l = group['labels'][i] if data_i == 0 else ''
if i < 6 and 'contains_base' in group and group[
'contains_base']:
ls = 'dashed'
else:
ls = '-'
dashes = () # type: Tuple
if idf.opt['plotErrors']:
if idf.opt['plotPrioriTorques']:
n = 3
else:
n = 2
if i == n:
ls = 'dashed'
dashes = (3, 0.5)
ax.plot(d['time'][::skip],
d['data'][data_i][::skip, i],
label=l,
color=colors[i],
alpha=1 - (data_i / 2.0),
linestyle=ls,
dashes=dashes)
else:
#data vector
ax.plot(d['time'][::skip],
d['data'][data_i][::skip],
label=group['labels'][d_i],
color=colors[0],
alpha=1 - (data_i / 2.0))
ax.grid(which='both', linestyle="dotted", alpha=0.8)
if 'y_label' in group:
ax.set_ylabel(group['y_label'])
if idf.opt['outputAs'] != 'tikz':
ax.set_xlabel("Time (s)")
plt.setp([a.get_xticklabels() for a in axes[:-1]], visible=False)
#plt.setp([a.get_yticklabels() for a in axes], fontsize=8)
if idf.opt['plotLegend']:
handles, labels = ax.get_legend_handles_labels()
if idf.opt['outputAs'] == 'html':
#TODO: show legend properly (see mpld3 bug #274)
#leg = fig.legend(handles, labels, loc='upper right', fancybox=True, fontsize=10, title='')
leg = axes[0].legend(handles,
labels,
loc='upper right',
fancybox=True,
fontsize=10,
title='',
prop={'size': 8})
else:
leg = plt.figlegend(handles,
labels,
loc='upper right',
fancybox=True,
fontsize=font_size,
title='',
prop={'size': font_size - 3})
leg.draggable()
fig.subplots_adjust(hspace=2)
fig.set_tight_layout(True)
if idf.opt['outputAs'] == 'html':
plugins.clear(fig)
plugins.connect(fig, plugins.Reset(), plugins.BoxZoom(),
plugins.Zoom(enabled=False),
plugins.MousePosition(fontsize=14, fmt=".5g"))
figures.append(mpld3.fig_to_html(fig))
elif idf.opt['outputAs'] == 'interactive':
plt.show(block=False)
elif idf.opt['outputAs'] == 'pdf':
pp.savefig(plt.gcf())
elif idf.opt['outputAs'] == 'tikz':
from matplotlib2tikz import save as tikz_save
tikz_save('{}_{}_{}.tex'.format(
filename, group['dataset'][0]['title'].replace('_', '-'),
ds // idf.model.num_dofs),
figureheight='\\figureheight',
figurewidth='\\figurewidth',
show_info=False)
if idf.opt['outputAs'] == 'html':
path = os.path.dirname(os.path.abspath(__file__))
template_environment = Environment(
autoescape=False,
loader=FileSystemLoader(os.path.join(path, '../output')),
trim_blocks=False)
context = {'figures': figures, 'text': self.text}
outfile = os.path.join(path, '..', 'output', filename)
import codecs
with codecs.open(outfile, 'w', 'utf-8') as f:
html = template_environment.get_template(
"templates/index.html").render(context)
f.write(html)
print("Saved output at file://{}".format(outfile))
elif idf.opt['outputAs'] == 'interactive':
#keep non-blocking plot windows open
plt.show()
elif idf.opt['outputAs'] == 'pdf':
pp.close()