def _organise_data(self, conf):
arg = self._extractor.arg()
arg["country_codes"] = conf.countries
arg["indicator_codes"] = conf.indicators
#arg["rca_indicator_codes"] = conf.rcaIndicators
arg["interval"] = (conf.start_date, conf.end_date)
countries = self._extractor.grab(arg)
#TODO: add the process method somewhere inside the preprocessor
#self._extractor.process(conf.process_indicators,
# method = "slope",
# look_back_years=conf.look_back_years)
print("organiser got back:")
print(countries)
values = []
for country in countries:
years = range(conf.start_date, conf.end_date)
crisis_seer = CrisisSeer(conf.sample_selection_file)
crisis_years = crisis_seer.get_crisis_years(country.code)
all_x = []
for t in years:
x = {}
x['country'] = country.code
x['date'] = t
x['crisis'] = t in crisis_years
emptyIndicators = 0
for indicator_code in country.indicator_codes:
indicator = country.get_indicator(indicator_code)
try:
x[indicator_code] = indicator.get_value_at(t)
except NonExistentDataError:
x[indicator_code] = ""
emptyIndicators = emptyIndicators + 1
if (len(country.indicator_codes) == emptyIndicators): continue
all_x.append(x)
values = values + all_x
self.vis_data = {
'countries': conf.countries,
'indicators': conf.indicators,
'values': values
}
return self.vis_data
def _organise_data(self, conf):
arg = self._extractor.arg()
arg["country_codes"] = conf.countries
arg["indicator_codes"] = conf.indicators
arg["interval"] = (conf.start_date, conf.end_date)
countries = self._extractor.grab(arg)
#TODO: add the process method somewhere inside the preprocessor
#self._extractor.process(conf.process_indicators,
# method = "slope",
# look_back_years=conf.look_back_years)
self.vis_data = []
for country in countries:
#x_ind_code,y_ind_code = country.indicator_codes()[:2] # Throws error: 'list' object is not callable
x_ind_code,y_ind_code = country.indicator_codes[:2]
x_ind = country.get_indicator(x_ind_code)
y_ind = country.get_indicator(y_ind_code)
years = range(conf.start_date, conf.end_date)
# arrange x and y vectors
all_x = []
all_x_dates = []
all_y = []
all_y_dates = []
for t in years:
try:
x = x_ind.get_value_at(t)
y = y_ind.get_value_at(t)
all_x.append(x)
all_x_dates.append(t)
all_y.append(y)
all_y_dates.append(t)
except NonExistentDataError:
continue
crisis_seer = CrisisSeer(conf.sample_selection_file)
crisis_years = crisis_seer.get_crisis_years(country.code)
#'crises': crisis_years,
country_repr = {'code': country.code, 'dates': all_x_dates, 'crises': crisis_years,
#'x_ind': x_ind.simple_dict_repr(),
#'y_ind': y_ind.simple_dict_repr()}
'x_ind': {'code': x_ind.code, 'data': all_x},
'y_ind': {'code': y_ind.code, 'data': all_y}}
self.vis_data.append(country_repr)
return self.vis_data
def _organise_data(self, conf):
arg = self._extractor.arg()
arg["country_codes"] = conf.countries
arg["indicator_codes"] = conf.indicators
#arg["rca_indicator_codes"] = conf.rcaIndicators
arg["interval"] = (conf.start_date, conf.end_date)
countries = self._extractor.grab(arg)
#TODO: add the process method somewhere inside the preprocessor
#self._extractor.process(conf.process_indicators,
# method = "slope",
# look_back_years=conf.look_back_years)
print("organiser got back:")
print(countries)
values = []
for country in countries:
years = range(conf.start_date, conf.end_date)
crisis_seer = CrisisSeer(conf.sample_selection_file)
crisis_years = crisis_seer.get_crisis_years(country.code)
all_x = []
for t in years:
x = {}
x['country'] = country.code
x['date'] = t
x['crisis'] = t in crisis_years
emptyIndicators = 0
for indicator_code in country.indicator_codes:
indicator = country.get_indicator(indicator_code)
try:
x[indicator_code] = indicator.get_value_at(t)
except NonExistentDataError:
x[indicator_code] = ""
emptyIndicators = emptyIndicators + 1
if (len(country.indicator_codes)==emptyIndicators): continue
all_x.append(x)
values = values + all_x
self.vis_data = {'countries': conf.countries, 'indicators': conf.indicators, 'values': values}
return self.vis_data
def _create_figure(self, item):
"""
_create_figure a time series marked with crisis and model data
for the data defined in the conf file
"""
crisis_seer = CrisisSeer(conf.sample_selection_file)
i = 0
# we didn't override _get_items, so we will get a country
country = item
#crisis_years = set(conf.manual_crises[i])
crisis_years = crisis_seer.get_crisis_years(country.code)
i = i + 1
x_ind_code, y_ind_code = country.indicator_codes[:2]
x_ind = country.get_indicator(x_ind_code)
y_ind = country.get_indicator(y_ind_code)
years = range(conf.start_date, conf.end_date)
all_x = []
all_x_dates = []
all_y = []
all_y_dates = []
if self._should_add_meta_marks():
labeled_crisis, labeled_model, labeled_outside_model = False, False, False
else:
labeled_crisis, labeled_model, labeled_outside_model = True, True, True
for t in years:
try:
x = x_ind.get_value_at(t)
y = y_ind.get_value_at(t)
all_x.append(x)
all_x_dates.append(t)
all_y.append(y)
all_y_dates.append(t)
xlabel(conf.indicator_titles[x_ind_code])
ylabel(conf.indicator_titles[y_ind_code])
legend_label = None
if t in crisis_years: # crisis
label_text = "Crisis: " + str(t)
if conf.dist_pixels:
dist_x = conf.label_dist_x
dist_y = conf.label_dist_y
else:
dist = math.sqrt(
math.pow(x - x_ind.get_value_at(t - 1), 2) +
math.pow(y - y_ind.get_value_at(t - 1), 2))
if conf.dist_2d:
dist_x = dist * conf.label_dist_factor_x
dist_y = dist * conf.label_dist_factor_y
else:
dist_y = dist_x = dist * conf.label_dist_factor_x
annotate(label_text,
size="medium",
xy=(x + dist_x, y + dist_y))
mark = conf.crisis_mark
colour = conf.crisis_colour
size = conf.crisis_size
if not labeled_crisis:
legend_label, labeled_crisis = "crisis", True
else:
if len(all_x) > conf.look_back_years and conf.model(
all_x, all_x_dates, all_y, all_y_dates):
#if conf.model(x,y):
# model true
colour = conf.model_true_colour
mark = conf.model_true_mark
size = conf.model_true_size
if not labeled_model:
legend_label, labeled_model = "satisfies model", True
else:
# model false
colour = conf.model_false_colour
mark = conf.model_false_mark
size = conf.model_false_size
if not labeled_outside_model:
legend_label, labeled_outside_model = "doesn't satisfy model", True
scatter(x,
y,
s=size,
c=colour,
alpha=1.0,
marker=mark,
label=legend_label)
except NonExistentDataError:
pass
if len(all_x) != len(
all_y): # there is no data for one of the indicators
print("Error: number of data points to be plotted not equal")
raise NonExistentDataError
try:
plot(all_x, all_y, c="black")
except ValueError:
print "Something wrong with data to be plotted"
raise
return self.figure
def _create_figure(self, item):
"""
_create_figure a time series marked with crisis and model data
for the data defined in the conf file
"""
crisis_seer = CrisisSeer(conf.sample_selection_file)
i = 0
# we didn't override _get_items, so we will get a country
country = item
#crisis_years = set(conf.manual_crises[i])
crisis_years = crisis_seer.get_crisis_years(country.code)
i = i +1
x_ind_code,y_ind_code = country.indicator_codes[:2]
x_ind = country.get_indicator(x_ind_code)
y_ind = country.get_indicator(y_ind_code)
years = range(conf.start_date, conf.end_date)
all_x = []
all_x_dates = []
all_y = []
all_y_dates = []
if self._should_add_meta_marks():
labeled_crisis, labeled_model, labeled_outside_model = False, False, False
else:
labeled_crisis, labeled_model, labeled_outside_model = True, True, True
for t in years:
try:
x = x_ind.get_value_at(t)
y = y_ind.get_value_at(t)
all_x.append(x)
all_x_dates.append(t)
all_y.append(y)
all_y_dates.append(t)
xlabel(conf.indicator_titles[x_ind_code])
ylabel(conf.indicator_titles[y_ind_code])
legend_label = None
if t in crisis_years: # crisis
label_text = "Crisis: " + str(t)
if conf.dist_pixels:
dist_x = conf.label_dist_x
dist_y = conf.label_dist_y
else:
dist = math.sqrt(math.pow(x-x_ind.get_value_at(t-1),2)
+math.pow(y-y_ind.get_value_at(t-1),2))
if conf.dist_2d:
dist_x = dist*conf.label_dist_factor_x
dist_y = dist*conf.label_dist_factor_y
else:
dist_y = dist_x = dist*conf.label_dist_factor_x
annotate(label_text, size="medium" , xy=(x+dist_x,y+dist_y))
mark = conf.crisis_mark
colour = conf.crisis_colour
size = conf.crisis_size
if not labeled_crisis: legend_label, labeled_crisis = "crisis", True
else:
if len(all_x)>conf.look_back_years and conf.model(all_x, all_x_dates, all_y, all_y_dates):
#if conf.model(x,y):
# model true
colour = conf.model_true_colour
mark = conf.model_true_mark
size = conf.model_true_size
if not labeled_model: legend_label, labeled_model = "satisfies model", True
else:
# model false
colour = conf.model_false_colour
mark = conf.model_false_mark
size = conf.model_false_size
if not labeled_outside_model: legend_label, labeled_outside_model = "doesn't satisfy model", True
scatter(x, y, s=size, c=colour,
alpha=1.0, marker=mark, label = legend_label)
except NonExistentDataError:
pass
if len(all_x)!=len(all_y): # there is no data for one of the indicators
print("Error: number of data points to be plotted not equal")
raise NonExistentDataError
try:
plot(all_x,all_y, c="black")
except ValueError:
print "Something wrong with data to be plotted"
raise
return self.figure
