def _process_and_display_trends(self, query, objects_expected):
from qtools.lib.nstats import moving_average_by_interval
from numpy import histogram
import numpy as np
group_by_plate = self.form_result['group_by_plate']
if objects_expected:
results = execute_built_query(query, group_by_plate, True, self.form_result['metric'][1])
else:
results = execute_built_query(query, group_by_plate, False)
epoch_results = [(float(stat), time.mktime(dt.timetuple()), id, name) for stat, dt, id, name in results if stat is not None]
exclude_func = create_exclude_function(self.form_result)
filtered_results = [tup for tup in epoch_results if exclude_func(tup[0])]
c.yaxis_title = fl.comparable_metric_display(MetricPattern.from_python(self.form_result['metric']))
c.mean_value = 'N/A'
c.std_value = 'N/A'
c.metric_name = fl.comparable_metric_display(MetricPattern.from_python(self.form_result['metric']))
c.table_data =[]
if group_by_plate:
moving_average = 10
epoch_url_results = [(stat, stamp*1000, url(controller='plate', action='view', id=id), name) for stat, stamp, id, name in filtered_results]
else:
moving_average = 50
epoch_url_results = [(stat, stamp*1000, url(controller='well', action='view', id=id), name) for stat, stamp, id, name in filtered_results]
if epoch_url_results:
c.stats = {}
stats, time_points, urls, names = zip(*epoch_url_results)
c.stats = h.literal(json.dumps(zip(time_points, stats)))
c.table_data = zip( names, stats, time_points )
#update if we can.
c.mean_value = str( np.mean( stats ) )
c.std_value = str( np.std( stats ) )
INTERVAL = 24*60*60*1000
ma = moving_average_by_interval(stats, time_points, INTERVAL, 10)
c.moving_avg = h.literal(json.dumps(ma))
c.urls = h.literal(json.dumps(urls))
c.names = h.literal(json.dumps(names))
# get histogram of points by day
#raise Exception, len(ma)
density, edges = histogram(time_points, bins=len(ma), range=(ma[0][0], ma[-1][0]))
c.densities = h.literal(json.dumps(zip([m[0] for m in ma], [int(d) for d in density])))
c.max_density_axis = 3*max(density)
else:
c.stats = {}
c.moving_avg = {}
c.urls = {}
c.names = {}
c.densities = {}
c.max_density_axis = 3
return render('/trend/results.html')
def _process_and_display_qcc(self, query, objects_expected):
"""
This procuces the data for java scirpt display of qcc
"""
import numpy as np
#average results across plates?
group_by_plate = self.form_result['group_by_plate']
if objects_expected:
results = execute_built_query(query, group_by_plate, True, self.form_result['metric'][1])
else:
results = execute_built_query(query, group_by_plate, False)
epoch_results = [(float(stat), time.mktime(dt.timetuple()), id, name, DR) for stat, dt, id, name, DR in results if stat is not None]
exclude_func = create_exclude_function(self.form_result)
filtered_results = [tup for tup in epoch_results if exclude_func(tup[0])]
c.yaxis_title = fl.comparable_metric_display(MetricPattern.from_python(self.form_result['metric']))
c.y_label = fl.comparable_metric_display(MetricPattern.from_python(self.form_result['metric']))
c.explanation, c.x_label = self._get_qcc_text()
if group_by_plate:
epoch_url_results = [];
for i in range( len( filtered_results ) ):
stat, stamp, id, name, DR = filtered_results[i]
url_str = url(controller='plate', action='view', id=id)
epoch_url_results.append( (stat,i,url_str,name,DR) )
#epoch_url_results = [(stat, stamp, url(controller='plate', action='view', id=id), name) for stat, stamp, id, name in filtered_results]
else:
epoch_url_results = [];
for i in range( len( filtered_results ) ):
stat, stamp, id, name, plate = filtered_results[i]
url_str = url(controller='well', action='view', id=id)
epoch_url_results.append( (stat,i,url_str,name,plate) )
#epoch_url_results = [(stat, stamp, url(controller='well', action='view', id=id), name) for stat, stamp, id, name in filtered_results]
if epoch_url_results:
c.stats = {}
stats, time_points, urls, names, xTickLabels = zip(*epoch_url_results)
xTickLabels = [ x + " " for x in xTickLabels]
c.stats = h.literal(json.dumps(zip(time_points, stats)))
stat_mean = np.mean( stats )
stat_standard_dev = np.std( stats )
#stat_standard_error = np.std( stats ) / np.sqrt( len(stats) )
ucl = stat_mean + 3 * stat_standard_dev
lcl = stat_mean - 3 * stat_standard_dev
mv = [[epoch_url_results[0][1],stat_mean],[epoch_url_results[-1][1],stat_mean]]
c.mean_value = h.literal(json.dumps(mv))
xaxis_start = epoch_url_results[0][1]
xaxis_end = epoch_url_results[-1][1]
c.ucl = h.literal(json.dumps([[xaxis_start,ucl],[xaxis_end,ucl]] ))
c.lcl = h.literal(json.dumps([[xaxis_start,lcl],[xaxis_end,lcl]] ))
axis_offset = 0.05*(max(stats)-min(stats))
if self.form_result['lower_yaxis']:
c.min_y_axis = float( self.form_result['lower_yaxis'] )
else:
c.min_y_axis = min( stats ) - axis_offset
if self.form_result['upper_yaxis']:
c.max_y_axis = float( self.form_result['upper_yaxis'] )
else:
c.max_y_axis = max( stats ) + axis_offset
c.mv_label = "Center Line (mean): %f" % stat_mean
c.ucl_label = "UCL (mean + 3std): %f" % ucl
c.lcl_label = "LCL (mean - 3std): %f" % lcl
# add user defined limits if presented
if self.form_result['upper_spec']:
upper_spec = float(self.form_result['upper_spec'])
c.upper_spec = h.literal(json.dumps([[xaxis_start,upper_spec],[xaxis_end,upper_spec]] ))
c.upper_spec_label = 'User defined upper line: %f' % upper_spec
else:
c.upper_spec = 'null'
c.upper_spec_label = ''
if self.form_result['lower_spec']:
lower_spec = float(self.form_result['lower_spec'])
c.lower_spec = h.literal(json.dumps([[xaxis_start,lower_spec],[xaxis_end,lower_spec]] ))
c.lower_spec_label = 'User defined lower line: %f' % lower_spec
else:
c.lower_spec = 'null'
c.lower_spec_label = ''
c.urls = h.literal(json.dumps(urls))
c.names = h.literal(json.dumps(names))
c.TickData = h.literal(json.dumps(zip(range(len(xTickLabels)),xTickLabels)))
else:
c.stats = {}
c.mean_value = {}
c.ucl = {}
c.lcl = {}
c.mv_label = ''
c.ucl_label = ''
c.lcl_label = ''
c.urls = {}
c.names = {}
c.TickData = {}
c.min_y_axis = 0
c.max_y_axis = 3
c.upper_spec = 'null'
c.upper_spec_label = ''
c.lower_spec = 'null'
c.lower_spec_label = ''
return render('/qcCharts/results.html')