def execute_normalization(self,analysis_id_I,concentration_units_I=[],plot_values_I = False,r_calc_I=None):
'''glog normalize concentration values using R'''
print('execute_normalization...')
if r_calc_I: r_calc = r_calc_I;
else: r_calc = r_calculate();
mplot = matplot();
# get the analysis information
analysis_info = [];
analysis_info = self.get_rows_analysisID_dataPreProcessingAnalysis(analysis_id_I);
# query metabolomics data from the experiment
data_transformed = [];
if concentration_units_I:
concentration_units = concentration_units_I;
else:
concentration_units = [];
for row in analysis_info:
concentration_units_tmp = []
concentration_units_tmp = self.get_concentrationUnits_experimentID_dataPreProcessingImputation(row['experiment_id']);
concentration_units.extend(concentration_units_tmp)
concentration_units = list(set(concentration_units));
for cu in concentration_units:
print('calculating normalization for concentration_units ' + cu);
data = [];
# get all of the samples in the simulation
for row in analysis_info:
data_tmp = [];
data_tmp = self.get_RExpressionData_AnalysisIDAndExperimentIDAndTimePointAndUnitsAndSampleNameShort_dataPreProcessingImputation(analysis_id_I,row['experiment_id'], row['time_point'], cu, row['sample_name_short']);
data.extend(data_tmp)
# call R
concentrations = None;
concentrations_glog = None;
data_glog, concentrations, concentrations_glog = r_calc.calculate_normalization(data)
for i,d in enumerate(data_glog):
data_glog[i]['calculated_concentration_units']=cu+'_glog_normalization'
# plot original values:
if plot_values_I:
mplot.densityPlot(concentrations);
mplot.densityPlot(concentrations_glog);
# upload data
for d in data:
row = None;
row = data_dataPreProcessing_glogNormalization(analysis_id_I,d['experiment_id'], d['sample_name_short'],
d['time_point'],d['component_group_name'],
d['component_name'],d['calculated_concentration'],
d['calculated_concentration_units'] + '_glog_normalization',
True,None);
self.session.add(row);
data_transformed.extend(data_glog);
# commit data to the session every timepoint
self.session.commit();
self.update_concentrations_dataPreProcessingNormalization(analysis_id_I, data_transformed)
def export_boxAndWhiskersPlot_peakInformation_matplot(self,experiment_id_I,
peakInfo_parameter_I = ['height','retention_time','width_at_50','signal_2_noise'],
component_names_I=[],
filename_O = 'tmp',
figure_format_O = '.png'):
'''generate a boxAndWhiskers plot from peakInformation table'''
#TODO: remove after refactor
mplot = matplot();
print('export_boxAndWhiskersPlot...')
if peakInfo_parameter_I:
peakInfo_parameter = peakInfo_parameter_I;
else:
peakInfo_parameter = [];
peakInfo_parameter = self.get_peakInfoParameter_experimentID_dataStage01PeakInformation(experiment_id_I);
for parameter in peakInfo_parameter:
data_plot_mean = [];
data_plot_cv = [];
data_plot_ci = [];
data_plot_parameters = [];
data_plot_component_names = [];
data_plot_data = [];
data_plot_units = [];
if component_names_I:
component_names = component_names_I;
else:
component_names = [];
component_names = self.get_componentNames_experimentIDAndPeakInfoParameter_dataStage01PeakInformation(experiment_id_I,parameter);
for cn in component_names:
print('generating boxAndWhiskersPlot for component_name ' + cn);
# get the data
data = {};
data = self.get_row_experimentIDAndPeakInfoParameterComponentName_dataStage01PeakInformation(experiment_id_I,parameter,cn)
if data and data['peakInfo_ave']:
# record data for plotting
data_plot_mean.append(data['peakInfo_ave']);
data_plot_cv.append(data['peakInfo_cv']);
data_plot_ci.append([data['peakInfo_lb'],data['peakInfo_ub']]);
data_plot_data.append(data['peakInfo_data']);
data_plot_parameters.append(parameter);
data_plot_component_names.append(data['component_group_name']);
data_plot_units.append('Retention_time [min]');
# visualize the stats:
data_plot_se = [(x[1]-x[0])/2 for x in data_plot_ci]
filename = filename_O + '_' + experiment_id_I + '_' + parameter + figure_format_O;
mplot.boxAndWhiskersPlot(data_plot_parameters[0],data_plot_component_names,data_plot_units[0],'samples',data_plot_data,data_plot_mean,data_plot_ci,filename_I=filename,show_plot_I=False);
def export_boxAndWhiskersPlot_physiologicalRatios_matplot(self,experiment_id_I,sample_name_abbreviations_I=[],ratio_ids_I=[]):
'''generate a boxAndWhiskers plot from physiological ratios table'''
mplot = matplot();
print('execute_boxAndWhiskersPlot...')
# get time points
time_points = [];
time_points = self.get_timePoint_experimentID_dataStage01PhysiologicalRatiosAverages(experiment_id_I);
for tp in time_points:
print('generating boxAndWhiskersPlot for time_point ' + tp);
if ratio_ids_I:
ratio_ids = ratio_ids_I;
else:
ratio_ids = list(self.ratios.keys());
for k in ratio_ids:
#for k,v in self.ratios.iteritems():
print('generating boxAndWhiskersPlot for ratio ' + k); # get sample_name_abbreviations
data_plot_mean = [];
data_plot_var = [];
data_plot_ci = [];
data_plot_sna = [];
data_plot_ratio_ids = [];
data_plot_data = [];
data_plot_ratio_units = [];
if sample_name_abbreviations_I:
sample_name_abbreviations = sample_name_abbreviations_I;
else:
sample_name_abbreviations = [];
sample_name_abbreviations = self.get_sampleNameAbbreviations_experimentIDAndTimePointAndRatioID_dataStage01PhysiologicalRatiosAverages(experiment_id_I,tp,k);
for sna in sample_name_abbreviations:
print('generating boxAndWhiskersPlot for sample_name_abbreviation ' + sna);
# get the data
data = {};
data = self.get_data_experimentIDAndTimePointAndRatioIDAndSampleNameAbbreviation_dataStage01PhysiologicalRatiosAverages(experiment_id_I,tp,k,sna)
ratio_values = [];
ratio_values = self.get_ratios_experimentIDAndSampleNameAbbreviationAndTimePointAndRatioID_dataStage01PhysiologicalRatiosReplicates(experiment_id_I,sna,tp,k)
# record data for plotting
data_plot_mean.append(data['physiologicalratio_value_ave']);
data_plot_var.append(data['physiologicalratio_value_cv']);
data_plot_ci.append([data['physiologicalratio_value_lb'],data['physiologicalratio_value_ub']]);
data_plot_data.append(ratio_values);
data_plot_sna.append(sna);
data_plot_ratio_ids.append(k);
data_plot_ratio_units.append('');
# visualize the stats:
#self.matplot.barPlot(data_plot_ratio_ids[0],data_plot_sna,data_plot_sna[0],'samples',data_plot_mean,data_plot_var);
mplot.boxAndWhiskersPlot(data_plot_ratio_ids[0],data_plot_sna,data_plot_ratio_units[0],'samples',data_plot_data,data_plot_mean,data_plot_ci);
def export_scatterLinePlot_physiologicalRatios_matplot(self,experiment_id_I,sample_name_abbreviations_I=[],ratio_ids_I=[]):
'''Generate a scatter line plot for physiological ratios averages'''
mplot = matplot();
print('Generating scatterLinePlot for physiologicalRatios')
# get time points
time_points = [];
time_points = self.get_timePoint_experimentID_dataStage01PhysiologicalRatiosReplicates(experiment_id_I);
for tp in time_points:
print('Generating scatterLinePlot for physiologicalRatios for time_point ' + tp);
# get physiological ratio_ids
ratios = {};
ratios = self.get_ratioIDs_experimentIDAndTimePoint_dataStage01PhysiologicalRatiosReplicates(experiment_id_I,tp);
for k,v in ratios.items():
if ratio_ids_I:
if not k in ratio_ids_I:
continue;
print('Generating scatterLinePlot for physiologicalRatios for ratio ' + k);
# get sample_names
if sample_name_abbreviations_I:
sample_name_abbreviations = sample_name_abbreviations_I;
else:
sample_name_abbreviations = [];
sample_name_abbreviations = self.get_sampleNameAbbreviations_experimentIDAndTimePointAndRatioID_dataStage01PhysiologicalRatiosAverages(experiment_id_I,tp,k);
ratios_num = [];
ratios_den = [];
for sna_cnt,sna in enumerate(sample_name_abbreviations):
print('Generating scatterLinePlot for physiologicalRatios for sample name abbreviation ' + sna);
# get ratios_numerator
ratio_num = None;
ratio_num = self.get_ratio_experimentIDAndTimePointAndRatioIDAndSampleNameAbbreviation_dataStage01PhysiologicalRatiosAverages(experiment_id_I,tp,k+'_numerator',sna)
if not ratio_num: continue;
ratios_num.append(ratio_num);
# get ratios_denominator
ratio_den = None;
ratio_den = self.get_ratio_experimentIDAndTimePointAndRatioIDAndSampleNameAbbreviation_dataStage01PhysiologicalRatiosAverages(experiment_id_I,tp,k+'_denominator',sna)
if not ratio_den: continue;
ratios_den.append(ratio_den);
# plot the data
mplot.scatterLinePlot(k,k+'_denominator',k+'_numerator',ratios_den,ratios_num,sample_name_abbreviations);
def plot_averageSpectrumNormSum(self,experiment_id_I, time_points_I = None, sample_name_abbreviations_I = None, met_ids_I = None, scan_types_I = None):
'''calculate the average normalized intensity for all samples and scan types'''
'''Assumptions:
only a single fragment:spectrum is used_ per sample name abbreviation, time-point, replicate, scan_type
(i.e. there are no multiple dilutions of the same precursor:spectrum that are used_)
'''
mids = mass_isotopomer_distributions();
print('plot_averagesNormSum...')
plot = matplot();
# get time points
if time_points_I:
time_points = time_points_I;
else:
time_points = [];
time_points = self.get_timePoint_experimentID_dataStage01AveragesNormSum(experiment_id_I);
for tp in time_points:
print('Plotting product and precursor for time-point ' + str(tp));
# get sample names and sample name abbreviations
if sample_name_abbreviations_I:
sample_abbreviations = sample_name_abbreviations_I;
sample_types_lst = ['Unknown' for x in sample_abbreviations];
else:
sample_abbreviations = [];
sample_types = ['Unknown'];
sample_types_lst = [];
for st in sample_types:
sample_abbreviations_tmp = [];
sample_abbreviations_tmp = self.get_sampleNameAbbreviations_experimentIDAndSampleTypeAndTimePoint_dataStage01AveragesNormSum(experiment_id_I,st,tp);
sample_abbreviations.extend(sample_abbreviations_tmp);
sample_types_lst.extend([st for i in range(len(sample_abbreviations_tmp))]);
for sna_cnt,sna in enumerate(sample_abbreviations):
print('Plotting product and precursor for sample name abbreviation ' + sna);
# get the scan_types
if scan_types_I:
scan_types = [];
scan_types_tmp = [];
scan_types_tmp = self.get_scanTypes_experimentIDAndTimePointAndSampleAbbreviationsAndSampleType_dataStage01AveragesNormSum(experiment_id_I,tp,sna,sample_types_lst[sna_cnt]);
scan_types = [st for st in scan_types_tmp if st in scan_types_I];
else:
scan_types = [];
scan_types = self.get_scanTypes_experimentIDAndTimePointAndSampleAbbreviationsAndSampleType_dataStage01AveragesNormSum(experiment_id_I,tp,sna,sample_types_lst[sna_cnt]);
for scan_type in scan_types:
print('Plotting product and precursor for scan type ' + scan_type)
# met_ids
if not met_ids_I:
met_ids = [];
met_ids = self.get_metIDs_experimentIDAndSampleAbbreviationAndTimePointAndSampleTypeAndScanType_dataStage01AveragesNormSum( \
experiment_id_I,sna,tp,sample_types_lst[sna_cnt],scan_type);
else:
met_ids = met_ids_I;
if not(met_ids): continue #no component information was found
for met in met_ids:
print('Plotting product and precursor for metabolite ' + met);
# fragments
fragment_formulas = [];
fragment_formulas = self.get_fragmentFormula_experimentIDAndSampleAbbreviationAndTimePointAndSampleTypeAndScanTypeAndMetID_dataStage01AveragesNormSum( \
experiment_id_I,sna,tp,sample_types_lst[sna_cnt],scan_type,met);
for frag in fragment_formulas:
print('Plotting product and precursor for fragment ' + frag);
# data
data_mat = [];
data_mat_cv = [];
data_masses = [];
data_mat,data_mat_cv,data_masses = self.get_spectrum_experimentIDAndSampleAbbreviationAndTimePointAndSampleTypeAndScanTypeAndMetIDAndFragmentFormula_dataStage01AveragesNormSum( \
experiment_id_I,sna,tp,sample_types_lst[sna_cnt],scan_type,met,frag);
data_stdev = [];
for i,d in enumerate(data_mat):
stdev = 0.0;
stderr = 0.0;
if data_mat_cv[i]:
stdev = data_mat[i]*data_mat_cv[i]/100;
data_stdev.append(stdev);
title = sna+'_'+met+'_'+frag;
plot.barPlot(title,data_masses,'intensity','m/z',data_mat,var_I=None,se_I=data_stdev,add_labels_I=True)
def plot_normalizedSpectrumNormSum(self,experiment_id_I, sample_names_I = None, sample_name_abbreviations_I = None, met_ids_I = None, scan_types_I = None):
'''calculate the average normalized intensity for all samples and scan types'''
'''Assumptions:
only a single fragment:spectrum is used_ per sample name abbreviation, time-point, replicate, scan_type
(i.e. there are no multiple dilutions of the same precursor:spectrum that are used_)
'''
mids = mass_isotopomer_distributions();
print('plot_normalizedSpectrumNormSum...')
plot = matplot();
# get time points
time_points = self.get_timePoint_experimentID_dataStage01Normalized(experiment_id_I);
for tp in time_points:
print('Plotting precursor and product spectrum from isotopomer normalized for time-point ' + str(tp));
if sample_names_I:
sample_abbreviations = [];
sample_types = ['Unknown','QC'];
sample_types_lst = [];
for sn in sample_names_I:
for st in sample_types:
sample_abbreviations_tmp = [];
sample_abbreviations_tmp = self.get_sampleNameAbbreviations_experimentIDAndSampleTypeAndTimePointAndSampleName_dataStage01Normalized(experiment_id_I,st,tp,sn);
sample_abbreviations.extend(sample_abbreviations_tmp);
sample_types_lst.extend([st for i in range(len(sample_names_tmp))]);
elif sample_name_abbreviations_I:
sample_abbreviations = sample_name_abbreviations_I;
sample_types_lst = ['Unknown' for x in sample_abbreviations];
# query sample types from sample name abbreviations and time-point from data_stage01_isotopomer_normalized
else:
# get sample names and sample name abbreviations
sample_abbreviations = [];
sample_types = ['Unknown','QC'];
sample_types_lst = [];
for st in sample_types:
sample_abbreviations_tmp = [];
sample_abbreviations_tmp = self.get_sampleNameAbbreviations_experimentIDAndSampleTypeAndTimePoint_dataStage01Normalized(experiment_id_I,st,tp);
sample_abbreviations.extend(sample_abbreviations_tmp);
sample_types_lst.extend([st for i in range(len(sample_abbreviations_tmp))]);
for sna_cnt,sna in enumerate(sample_abbreviations):
print('Plotting precursor and product spectrum from isotopomer normalized for sample name abbreviation ' + sna);
# get the scan_types
if scan_types_I:
scan_types = [];
scan_types_tmp = [];
scan_types_tmp = self.get_scanTypes_experimentIDAndTimePointAndSampleAbbreviationsAndSampleType_dataStage01Normalized(experiment_id_I,tp,sna,sample_types_lst[sna_cnt]);
scan_types = [st for st in scan_types_tmp if st in scan_types_I];
else:
scan_types = [];
scan_types = self.get_scanTypes_experimentIDAndTimePointAndSampleAbbreviationsAndSampleType_dataStage01Normalized(experiment_id_I,tp,sna,sample_types_lst[sna_cnt]);
for scan_type in scan_types:
print('Plotting precursor and product spectrum for scan type ' + scan_type)
# met_ids
if not met_ids_I:
met_ids = [];
met_ids = self.get_metIDs_experimentIDAndSampleAbbreviationAndTimePointAndSampleTypeAndScanType_dataStage01Normalized( \
experiment_id_I,sna,tp,sample_types_lst[sna_cnt],scan_type);
else:
met_ids = met_ids_I;
if not(met_ids): continue #no component information was found
for met in met_ids:
print('Plotting precursor and product spectrum for metabolite ' + met);
# get replicates
replicate_numbers = [];
replicate_numbers = self.get_replicateNumbers_experimentIDAndSampleAbbreviationAndTimePointAndScanTypeAndMetID_dataStage01Normalized( \
experiment_id_I,sna,tp,scan_type,met);
peakSpectrum_normalized_lst = [];
for rep in replicate_numbers:
print('Plotting precursor and product spectrum for replicate_number ' + str(rep));
#get data
peakData_I = {};
peakData_I = self.get_dataNormalized_experimentIDAndSampleAbbreviationAndTimePointAndScanTypeAndMetIDAndReplicateNumber_dataStage01Normalized( \
experiment_id_I,sna,tp,scan_type,met,rep);
fragment_formulas = list(peakData_I.keys());
peakSpectrum_corrected, peakSpectrum_normalized = mids.extract_peakList_normSum(\
peakData_I, fragment_formulas, True);
peakSpectrum_normalized_lst.append(peakSpectrum_normalized);
# plot spectrum data for all replicates and fragments
fragment_formulas_unique = list(set(fragment_formulas_lst));
for fragment in fragment_formulas_unique:
panelLabels = [];
xticklabels = [];
mean = [];
xlabel = 'm/z'
ylabel = 'intensity'
for rep,spectrum in enumerate(peakSpectrum_normalized_lst):
panelLabels_tmp = sna+'_'+met+'_'+fragment+'_'+str(rep+1)
xticklabels_tmp = [];
mean_tmp = [];
for mass,intensity in spectrum[fragment].items():
intensity_tmp = intensity;
if not intensity_tmp: intensity_tmp=0.0
mean_tmp.append(intensity_tmp);
xticklabels_tmp.append(mass);
panelLabels.append(panelLabels_tmp);
xticklabels.append(xticklabels_tmp);
mean.append(mean_tmp);
plot.multiPanelBarPlot('',xticklabels,xlabel,ylabel,panelLabels,mean);
def export_scatterLinePlot_peakResolution_matplot(self,experiment_id_I,sample_names_I=[],sample_types_I=['Standard'],component_name_pairs_I=[],
peakInfo_I = ['rt_dif','resolution'],
acquisition_date_and_time_I=[None,None],
x_title_I='Time [hrs]',y_title_I='Retention Time [min]',y_data_type_I='acquisition_date_and_time',
plot_type_I='single'):
'''Analyze resolution for critical pairs'''
#Input:
# experiment_id_I
# sample_names_I
# sample_types_I
# component_name_pairs_I = [[component_name_1,component_name_2],...]
# acquisition_date_and_time_I = ['%m/%d/%Y %H:%M','%m/%d/%Y %H:%M']
#TODO: remove after refactor
mplot = matplot();
print('export_peakInformation_resolution...')
#convert string date time to datetime
# e.g. time.strptime('4/15/2014 15:51','%m/%d/%Y %H:%M')
acquisition_date_and_time = [];
if acquisition_date_and_time_I and acquisition_date_and_time_I[0] and acquisition_date_and_time_I[1]:
for dateandtime in acquisition_date_and_time_I:
time_struct = strptime(dateandtime,'%m/%d/%Y %H:%M')
dt = datetime.fromtimestamp(mktime(time_struct))
acquisition_date_and_time.append(dt);
else: acquisition_date_and_time=[None,None]
data_O = [];
component_names_pairs_all = [];
# get sample names
if sample_names_I and sample_types_I and len(sample_types_I)==1:
sample_names = sample_names_I;
sample_types = [sample_types_I[0] for sn in sample_names];
else:
sample_names = [];
sample_types = [];
for st in sample_types_I:
sample_names_tmp = [];
sample_names_tmp = self.get_sampleNames_experimentIDAndSampleType(experiment_id_I,st);
sample_names.extend(sample_names_tmp);
sample_types_tmp = [];
sample_types_tmp = [st for sn in sample_names_tmp];
sample_types.extend(sample_types_tmp);
for sn in sample_names:
print('analyzing peakInformation for sample_name ' + sn);
for component_name_pair in component_name_pairs_I:
# get critical pair data
cpd1 = {};
cpd2 = {};
cpd1 = self.get_peakInfo_sampleNameAndComponentName(sn,component_name_pair[0],acquisition_date_and_time);
cpd2 = self.get_peakInfo_sampleNameAndComponentName(sn,component_name_pair[1],acquisition_date_and_time);
# calculate the RT difference and resolution
rt_dif = 0.0;
rt_dif = abs(cpd1['retention_time']-cpd2['retention_time'])
resolution = 0.0;
resolution = rt_dif/(0.5*(cpd1['width_at_50']+cpd2['width_at_50']));
# record data
data_O.append({'component_name_pair':component_name_pair,
'rt_dif':rt_dif,
'resolution':resolution,
'component_group_name_pair':[cpd1['component_group_name'],cpd2['component_group_name']],
'sample_name':sn,
'acquisition_date_and_time':cpd1['acquisition_date_and_time']});
if plot_type_I == 'single':
for cnp in component_name_pairs_I:
data_parameters = {};
data_parameters_stats = {};
for parameter in peakInfo_I:
data_parameters[parameter] = [];
acquisition_date_and_times = [];
acquisition_date_and_times_hrs = [];
sample_names_parameter = [];
sample_types_parameter = [];
component_group_name_pair = None;
for sn_cnt,sn in enumerate(sample_names):
for d in data_O:
if d['sample_name'] == sn and d['component_name_pair'] == cnp and d[parameter]:
data_parameters[parameter].append(d[parameter]);
acquisition_date_and_times.append(d['acquisition_date_and_time'])
acquisition_date_and_times_hrs.append(d['acquisition_date_and_time'].year*8765.81277 + d['acquisition_date_and_time'].month*730.484 + d['acquisition_date_and_time'].day*365.242 + d['acquisition_date_and_time'].hour + d['acquisition_date_and_time'].minute / 60. + d['acquisition_date_and_time'].second / 3600.); #convert using datetime object
sample_names_parameter.append(sn);
sample_types_parameter.append(sample_types[sn_cnt])
component_group_name_pair = d['component_group_name_pair'];
# normalize time
acquisition_date_and_times_hrs.sort();
t_start = min(acquisition_date_and_times_hrs);
for t_cnt,t in enumerate(acquisition_date_and_times_hrs):
if y_data_type_I == 'acquisition_date_and_time':acquisition_date_and_times_hrs[t_cnt] = t - t_start;
elif y_data_type_I == 'count':acquisition_date_and_times_hrs[t_cnt] = t_cnt;
title = cn + '\n' + parameter;
filename = 'data/_output/' + experiment_id_I + '_' + cn + '_' + parameter + '.png'
mplot.scatterLinePlot(title,x_title_I,y_title_I,acquisition_date_and_times_hrs,data_parameters[parameter],fit_func_I='lowess',show_eqn_I=False,show_r2_I=False,filename_I=filename,show_plot_I=False);
if plot_type_I == 'multiple':
for parameter in peakInfo_I:
data_parameters = [];
acquisition_date_and_times = [];
acquisition_date_and_times_hrs = [];
sample_names_parameter = [];
sample_types_parameter = [];
component_group_names_pair = [];
component_names_pair = [];
for cnp_cnt,cnp in enumerate(component_name_pairs_I):
data = [];
acquisition_date_and_time = [];
acquisition_date_and_time_hrs = [];
sample_name_parameter = [];
sample_type_parameter = [];
for sn_cnt,sn in enumerate(sample_names):
for d in data_O:
if d['sample_name'] == sn and d['component_name_pair'] == cnp and d[parameter]:
data.append(d[parameter])
acquisition_date_and_time.append(d['acquisition_date_and_time'])
acquisition_date_and_time_hrs.append(d['acquisition_date_and_time'].year*8765.81277 + d['acquisition_date_and_time'].month*730.484 + d['acquisition_date_and_time'].day*365.242 + d['acquisition_date_and_time'].hour + d['acquisition_date_and_time'].minute / 60. + d['acquisition_date_and_time'].second / 3600.); #convert using datetime object
sample_name_parameter.append(sn);
sample_type_parameter.append(sample_types[sn_cnt])
if sn_cnt == 0:
component_group_names_pair.append(d['component_group_name_pair']);
component_names_pair.append(d['component_name_pair']);
# normalize time
acquisition_date_and_time_hrs.sort();
t_start = min(acquisition_date_and_time_hrs);
for t_cnt,t in enumerate(acquisition_date_and_time_hrs):
if y_data_type_I == 'acquisition_date_and_time':acquisition_date_and_time_hrs[t_cnt] = t - t_start;
elif y_data_type_I == 'count':acquisition_date_and_time_hrs[t_cnt] = t_cnt;
data_parameters.append(data);
acquisition_date_and_times.append(acquisition_date_and_time)
acquisition_date_and_times_hrs.append(acquisition_date_and_time_hrs);
sample_names_parameter.append(sample_name_parameter);
sample_types_parameter.append(sample_type_parameter)
# create data labels
data_labels = [];
for component_group_names in component_group_names_pair:
data_labels.append(component_group_names[0] + '/' + component_group_names[1]);
title = parameter;
filename = 'data/_output/' + experiment_id_I + '_' + parameter + '.eps'
mplot.multiScatterLinePlot(title,x_title_I,y_title_I,acquisition_date_and_times_hrs,data_parameters,data_labels_I=data_labels,fit_func_I=None,show_eqn_I=False,show_r2_I=False,filename_I=filename,show_plot_I=False);
def export_scatterLinePlot_peakInformation_matplot(self,experiment_id_I,sample_names_I=[],
sample_types_I=['Standard'],
component_names_I=[],
peakInfo_I = ['retention_time'],
acquisition_date_and_time_I=[None,None],
x_title_I='Time [hrs]',y_title_I='Retention Time [min]',y_data_type_I='acquisition_date_and_time',
plot_type_I='single',
filename_O = 'tmp',
figure_format_O = 'png'):
'''Analyze retention-time, height, s/n, and assymetry'''
#INPUT:
# experiment_id_I
# sample_names_I
# sample_types_I
# component_names_I
# peakInfo_I
# acquisition_date_and_time_I = ['%m/%d/%Y %H:%M','%m/%d/%Y %H:%M']
# y_data_type_I = 'acquisition_date_and_time' or 'count'
# plot_type_I = 'single', 'multiple', or 'sub'
print('export_peakInformation...')
#TODO: remove after refactor
mplot = matplot();
#convert string date time to datetime
# e.g. time.strptime('4/15/2014 15:51','%m/%d/%Y %H:%M')
acquisition_date_and_time = [];
if acquisition_date_and_time_I and acquisition_date_and_time_I[0] and acquisition_date_and_time_I[1]:
for dateandtime in acquisition_date_and_time_I:
time_struct = strptime(dateandtime,'%m/%d/%Y %H:%M')
dt = datetime.fromtimestamp(mktime(time_struct))
acquisition_date_and_time.append(dt);
else: acquisition_date_and_time=[None,None]
data_O = [];
component_names_all = [];
# get sample names
if sample_names_I and sample_types_I and len(sample_types_I)==1:
sample_names = sample_names_I;
sample_types = [sample_types_I[0] for sn in sample_names];
else:
sample_names = [];
sample_types = [];
for st in sample_types_I:
sample_names_tmp = [];
sample_names_tmp = self.get_sampleNames_experimentIDAndSampleType(experiment_id_I,st);
sample_names.extend(sample_names_tmp);
sample_types_tmp = [];
sample_types_tmp = [st for sn in sample_names_tmp];
sample_types.extend(sample_types_tmp);
for sn in sample_names:
print('analyzing peakInformation for sample_name ' + sn);
# get sample description
desc = {};
desc = self.get_description_experimentIDAndSampleID_sampleDescription(experiment_id_I,sn);
# get component names
if component_names_I:
component_names = component_names_I;
else:
component_names = [];
component_names = self.get_componentsNames_experimentIDAndSampleName(experiment_id_I,sn);
component_names_all.extend(component_names);
for cn in component_names:
# get rt, height, s/n
sst_data = {};
sst_data = self.get_peakInfo_sampleNameAndComponentName(sn,cn,acquisition_date_and_time);
if sst_data:
tmp = {};
tmp.update(sst_data);
tmp.update(desc);
tmp.update({'sample_name':sn});
data_O.append(tmp);
# Plot data over time
if component_names_I:
# use input order
component_names_unique = component_names_I;
else:
# use alphabetical order
component_names_unique = list(set(component_names_all));
component_names_unique.sort();
if plot_type_I == 'single':
for cn in component_names_unique:
data_parameters = {};
data_parameters_stats = {};
for parameter in peakInfo_I:
data_parameters[parameter] = [];
acquisition_date_and_times = [];
acquisition_date_and_times_hrs = [];
sample_names_parameter = [];
sample_types_parameter = [];
component_group_name = None;
for sn_cnt,sn in enumerate(sample_names):
for d in data_O:
if d['sample_name'] == sn and d['component_name'] == cn and d[parameter]:
data_parameters[parameter].append(d[parameter]);
acquisition_date_and_times.append(d['acquisition_date_and_time'])
acquisition_date_and_times_hrs.append(d['acquisition_date_and_time'].year*8765.81277 + d['acquisition_date_and_time'].month*730.484 + d['acquisition_date_and_time'].day*365.242 + d['acquisition_date_and_time'].hour + d['acquisition_date_and_time'].minute / 60. + d['acquisition_date_and_time'].second / 3600.); #convert using datetime object
sample_names_parameter.append(sn);
sample_types_parameter.append(sample_types[sn_cnt])
component_group_name = d['component_group_name'];
# normalize time
acquisition_date_and_times_hrs.sort();
t_start = min(acquisition_date_and_times_hrs);
for t_cnt,t in enumerate(acquisition_date_and_times_hrs):
if y_data_type_I == 'acquisition_date_and_time':acquisition_date_and_times_hrs[t_cnt] = t - t_start;
elif y_data_type_I == 'count':acquisition_date_and_times_hrs[t_cnt] = t_cnt;
title = cn + '\n' + parameter;
filename = filename_O + '_' + experiment_id_I + '_' + cn + '_' + parameter + figure_format_O;
mplot.scatterLinePlot(title,x_title_I,y_title_I,acquisition_date_and_times_hrs,data_parameters[parameter],fit_func_I='lowess',show_eqn_I=False,show_r2_I=False,filename_I=filename,show_plot_I=False);
if plot_type_I == 'multiple':
for parameter in peakInfo_I:
data_parameters = [];
acquisition_date_and_times = [];
acquisition_date_and_times_hrs = [];
sample_names_parameter = [];
sample_types_parameter = [];
component_group_names = [];
component_names = [];
for cn_cnt,cn in enumerate(component_names_unique):
data = [];
acquisition_date_and_time = [];
acquisition_date_and_time_hrs = [];
sample_name_parameter = [];
sample_type_parameter = [];
for sn_cnt,sn in enumerate(sample_names):
for d in data_O:
if d['sample_name'] == sn and d['component_name'] == cn and d[parameter]:
data.append(d[parameter])
acquisition_date_and_time.append(d['acquisition_date_and_time'])
acquisition_date_and_time_hrs.append(d['acquisition_date_and_time'].year*8765.81277 + d['acquisition_date_and_time'].month*730.484 + d['acquisition_date_and_time'].day*365.242 + d['acquisition_date_and_time'].hour + d['acquisition_date_and_time'].minute / 60. + d['acquisition_date_and_time'].second / 3600.); #convert using datetime object
sample_name_parameter.append(sn);
sample_type_parameter.append(sample_types[sn_cnt])
if sn_cnt == 0:
component_group_names.append(d['component_group_name']);
component_names.append(d['component_name']);
# normalize time
acquisition_date_and_time_hrs.sort();
t_start = min(acquisition_date_and_time_hrs);
for t_cnt,t in enumerate(acquisition_date_and_time_hrs):
if y_data_type_I == 'acquisition_date_and_time':acquisition_date_and_time_hrs[t_cnt] = t - t_start;
elif y_data_type_I == 'count':acquisition_date_and_time_hrs[t_cnt] = t_cnt;
data_parameters.append(data);
acquisition_date_and_times.append(acquisition_date_and_time)
acquisition_date_and_times_hrs.append(acquisition_date_and_time_hrs);
sample_names_parameter.append(sample_name_parameter);
sample_types_parameter.append(sample_type_parameter)
title = parameter;
filename = filename_O + '_' + experiment_id_I + '_' + parameter + figure_format_O;
mplot.multiScatterLinePlot(title,x_title_I,y_title_I,acquisition_date_and_times_hrs,data_parameters,data_labels_I=component_group_names,fit_func_I=None,show_eqn_I=False,show_r2_I=False,filename_I=filename,show_plot_I=False);