import pandas as pd import numpy as np import colorlover as cl import dash_html_components as html # TODO: Get these diretly from the google sheets #master_csv = 'assets/XJETS-master - Sheet1.csv' master_csv = 'assets/new_master_list.csv' col_attr = 'assets/XJETS-master - col_name_attributes_new.csv' data_availability = 'assets/data_downloads/data_info.csv' citations_file = 'assets/all_refs.csv' def generate_colorscale(values, colors): """ Generate a color scale for int type categories. Each category will be at the center of a bin. The will generate a discrete color bar but may have unequal bin widths """ max_val = np.max(values) min_val = np.min(values) if len(values) > len(colors): print('Length of values is greater than length of colors') return None nvals = len(values) if min_val != max_val: norm_values = [(i-min_val)/(max_val-min_val) for i in values] else: norm_values = [1] colorscale = [] ticklocations = [] for i in range(nvals): if i == 0: lb = norm_values[i] else: lb = prev_ub if i is not nvals-1: ub = (lb+norm_values[i+1])/2 else: ub = 1 prev_ub = ub colorscale.append([lb, colors[i]]) colorscale.append([ub, colors[i]]) ticklocations.append((lb+ub)*(max_val-min_val)/2) return colorscale, ticklocations class datastore: def __init__(self): try: self.master_df = pd.read_csv(master_csv) self.orig_columns=self.master_df.columns self.total_sources=len(self.master_df.index) self.col_names = self.master_df.columns self.refs_df = pd.read_csv(citations_file).fillna('') #print(self.col_names) self.data_availability = pd.read_csv(data_availability) self.master_df = self.master_df.merge( self.data_availability, how='left', on='Name') self.col_attr_df = pd.read_csv(col_attr).set_index('attribute').T self.state = 'good' self.endpoint_df = self.master_df.copy(deep=True) except ValueError as verr: print(verr) self.state = 'bad' def get_orig_columns(self): return self.orig_columns def get_master_data(self): return self.master_df def _filter_super_luminal_sources(self): self.endpoint_df = self.endpoint_df[self.endpoint_df['bapp3'] > 1] def _filter_xray_upstream_offset(self): self.endpoint_df = self.endpoint_df[self.endpoint_df['X-ray first'] > 0] def _filter_xray_downstream_offset(self): self.endpoint_df = self.endpoint_df[self.endpoint_df['Radio first'] > 0] def _filter_xray_upstream_downstream_offsets(self): self.endpoint_df = self.endpoint_df[ (self.endpoint_df['Radio first'] > 0) & ( self.endpoint_df['X-ray first'] > 0) ] def get_filtered_data(self): return self.endpoint_df def reset_filters(self): self.endpoint_df = self.master_df.copy(deep=True) def get_valid_scales(self): return self.col_attr_df[self.col_attr_df['use_as_scale'] == 1] def get_valid_categories(self): return self.col_attr_df[self.col_attr_df['use_as_cat'] == 1] def get_valid_axes(self): return self.col_attr_df[self.col_attr_df['valid_axis'].astype('int') == 1] def get_col_attributes(self): return self.col_attr_df def get_filters(self): return [ {'label': 'Super luminal sources', 'value': '_filter_super_luminal_sources'}, # {'label': 'Sources with upstream X-ray peak', # 'value': '_filter_xray_upstream_offset'}, # {'label': 'Sources with downstream X-ray peak', # 'value': '_filter_xray_downstream_offset'}, # {'label': 'Sources with both kind of offsets', # 'value': '_filter_xray_upstream_downstream_offsets'} ] def get_scaling_vars(self): scales_df = self.col_attr_df[self.col_attr_df['use_as_scale'].astype( 'int') == 1] options = [ { 'label': row['pretty_name'], 'value':i, 'primaryText':row['pretty_name'] } for i, row in scales_df.iterrows() ] options.insert(0,{ 'label': 'No scaling', 'value': None, 'primaryText': 'No scaling' }) return options def get_marker_scaling_options(self, scale_using): scales = self.endpoint_df[scale_using].fillna(0.1) max_derired_marker_size = 40 sizeref = 2*scales.max()/max_derired_marker_size**2 return { 'size': scales.tolist(), 'sizemode': 'area', 'sizemin': 4, 'sizeref': sizeref, 'color': None } def get_categorical_vars(self): cat_df = self.col_attr_df[self.col_attr_df['use_as_cat'].astype( 'int') == 1] options = [ { 'label': row['pretty_name'], 'value':i, 'primaryText':row['pretty_name'] } for i, row in cat_df.iterrows() ] options.insert(0,{'label': 'No Category', 'value': None, 'primaryText': 'No Category'}) return options def get_cat_color(self, cat_using): cat_values = self.endpoint_df[cat_using].fillna(0) # print(cat_values) # Use the divergent+spectral scale. # Assume that, the no. of sequential categories is less than 11 # if the variable is a continuous one interpolate the same if (cat_values % 1 == 0).all(): # sequential cat unique_vals = np.sort(cat_values.unique()) min_unique_val = np.min(unique_vals) n_unique_vals = len(unique_vals) if n_unique_vals < 1: return { 'color': None } elif n_unique_vals <= 3 and n_unique_vals > 0: colors = cl.scales['3']['qual']['Dark2'] else: colors = cl.scales[str(n_unique_vals)]['div']['Spectral'] colorscale, ticklocations = generate_colorscale( unique_vals, colors) # temporary hack # TODO: Fix the issue when there is only one unique value if(n_unique_vals == 1): ticklocations = [0] return { 'color': [i-min_unique_val for i in cat_values.tolist()], 'colorscale': colorscale, 'colorbar': { 'title': self.col_attr_df.loc[cat_using]['pretty_name'], 'titleside': 'top', 'tickmode': 'array', 'tickvals': ticklocations, 'ticktext': [str(int(i)) for i in unique_vals], } } else: return { 'color': None } def get_source_table(self, source): row = self.endpoint_df[self.endpoint_df['Name'] == source].iloc[0].fillna('NA') return html.Table( # [ # html.Tr( # [ # html.Th(self.col_attr_df.loc[col]['pretty_name']) for col in self.col_attr_df.index # ] # ) # ] # + [ html.Tr( [ html.Td(self.col_attr_df.loc[col]['pretty_name']), html.Td(row[col]) ] ) for col in self.col_names ], className='' ) def get_summary_data(self): return { 'total_resolved_xray_knots': self.master_df['#resolved X_ray knots'].sum(), 'total_knots_with_offsets': self.master_df['#knots with offsets'].sum(), 'total_knots_without_offsets': self.master_df['#knots without offsets'].sum(), 'total_resolved_xray_HS': self.master_df['#HS Xrays'].sum(), 'total_HS_with_offsets': self.master_df['#HS with offsets'].sum(), 'total_HS_without_offsets': self.master_df['#HS without offsets'].sum(), 'xray_first_offsets': self.master_df['X-ray first'].sum(), 'radio_first_offsets': self.master_df['Radio first'].sum(), 'total_xray_jet_terminates_at_a_bend': int(self.master_df['X-ray jet terminates before radio after a bend?'].sum()), 'total_jets':self.total_sources } def get_formatted_refs(self,source): print('src',source) refs=self.refs_df.query(f'Name2=="{source}"') #print(refs) if refs.shape[0]==0: return '' ref_list=[] def get_apa_cite(r): num='' if r['number']!='': num=f'({int(r["number"])})' return f"{r['author']} ({r['year']}), {r['journal2']}, {r['volume']}{num}, p.{r['pages']}" for index,row in refs.iterrows(): #print(row) ref_list.append(html.Li( [ get_apa_cite(row),html.Br(),f'"{row["title"]}"',html.Br(), html.A('[ADS]',href=row['link'],target="_blank") ] )) print('ref_l',ref_list) return html.Ol(ref_list)