import os import sys sys.path.insert(0, os.path.abspath(os.path.join(os.path.dirname(__file__), '..', '..'))) from module.config import cfg import matplotlib.pyplot as plt import numpy as np #------------------------------------------------------------------------------ # #------------------------------------------------------------------------------ def plot_sf_moments_pm(self, key, bounds, save = False, t_max=3011, ztf=False): """ Plots both positive and negative structure functions to investigate asymmetry in light curve. Parameters ---------- key : str Property from VAC bounds : array of z scores to use Returns ------- fig, ax, fig2, axes2, fig3, axes3: axes handles """ fig, ax = plt.subplots(1,1,figsize = (16,8)) fig2, axes2 = plt.subplots(2,1,figsize=(16,10)) fig3, axes3 = plt.subplots(8,1,figsize = (16,50)) label_range = {i:'{:.1f} < z < {:.1f}'.format(bounds[i],bounds[i+1]) for i in range(len(bounds)-1)} label_range_val = {i:'{:.1f} < {} < {:.1f}'.format(self.bounds_values[i],key,self.bounds_values[i+1]) for i in range(len(self.bounds_values)-1)} # label_moment = ['mean', 'std', 'skew_stand', 'Excess kurtosis'] label_moment = ['mean', 'Excess kurtosis'] cmap = plt.cm.jet for i in range(8): dms_binned, dts_binned, t_bin_edges, t_bin_chunk, t_bin_chunk_centres, m_bin_edges, m_bin_centres, t_dict = self.bin_dtdm(t_max = t_max, t_spacing='log', m_spacing='log', read_in=i+1, key=key, ztf=ztf) SF_n = (((m_bin_centres[:100]**2)*dms_binned[:,:100]).sum(axis=1)/dms_binned[:,:100].sum(axis=1))**0.5 SF_p = (((m_bin_centres[100:]**2)*dms_binned[:,100:]).sum(axis=1)/dms_binned[:,100:].sum(axis=1))**0.5 ax.plot(t_bin_chunk_centres, SF_p, label = label_range_val[i], lw = 0.5, marker = 'o', ls='-', color = cmap(i/10)) ax.plot(t_bin_chunk_centres, SF_n, label = label_range_val[i], lw = 0.5, marker = 'o', ls='--', color = cmap(i/10)) ax.legend() ax.set(yscale='log', xscale='log') axes3[i].hist(t_bin_chunk[:-1], weights = dms_binned[:,:100].sum(axis=1), alpha = 0.5, label = '-ve',bins = 19) axes3[i].hist(t_bin_chunk[:-1], weights = dms_binned[:,100:].sum(axis=1), alpha = 0.5, label = '+ve',bins = 19) axes3[i].set(yscale='log') dms_binned_norm = np.zeros((19,200)) moments = np.zeros(19) for j in range(19): dms_binned_norm[j],_= np.histogram(m_bin_edges[:-1], m_bin_edges, weights = dms_binned[j], density=True); # print('number of -ve dm in {} for {}: {:,}'.format(t_dict[j],label_range[i], dms_binned[j,:100].sum())) # print('number of +ve dm in {} for {}: {:,}'.format(t_dict[j],label_range[i], dms_binned[j,100:].sum())) moments = calc_moments(m_bin_centres,dms_binned_norm) for idx, ax2 in enumerate(axes2.ravel()): ax2.plot(t_bin_chunk_centres, moments[idx], lw = 0.5, marker = 'o', label = label_range_val[i], color = cmap(i/10.0)); # ax2.legend() ax2.set(xlabel='mjd', ylabel = label_moment[idx]) ax2.axhline(y=0, lw=0.5, ls = '--') ax.set(xlabel='mjd', ylabel = 'structure function') if save: # fig.savefig('SF_{}.pdf'.format(key),bbox_inches='tight') fig2.savefig('moments_{}.pdf'.format(key),bbox_inches='tight') return fig, ax, fig2, axes2, fig3, axes3 #------------------------------------------------------------------------------ # #------------------------------------------------------------------------------ def plot_sf_moments(self, key, bounds, save = False, t_max=3011, ztf=False): fig, ax = plt.subplots(1,1,figsize = (16,8)) fig2, axes2 = plt.subplots(2,1,figsize=(16,10)) fig3, axes3 = plt.subplots(8,1,figsize = (16,50)) label_range = {i:'{:.1f} < z < {:.1f}'.format(bounds[i],bounds[i+1]) for i in range(len(bounds)-1)} label_range_val = {i:'{:.1f} < {} < {:.1f}'.format(self.bounds_values[i],key,self.bounds_values[i+1]) for i in range(len(self.bounds_values)-1)} # label_moment = ['mean', 'std', 'skew_stand', 'Excess kurtosis'] label_moment = ['mean', 'Excess kurtosis'] cmap = plt.cm.jet for i in range(8): dms_binned, dts_binned, t_bin_edges, t_bin_chunk, t_bin_chunk_centres, m_bin_edges, m_bin_centres, t_dict = self.bin_dtdm(t_max = t_max, t_spacing='log', m_spacing='log', read_in=i+1, key=key, ztf=ztf) SF = (((m_bin_centres**2)*dms_binned).sum(axis=1)/dms_binned.sum(axis=1))**0.5 ax.plot(t_bin_chunk_centres, SF, label = label_range_val[i], lw = 0.5, marker = 'o', ls='-', color = cmap(i/10)) ax.legend() ax.set(yscale='log', xscale='log') axes3[i].hist(t_bin_chunk[:-1], weights = dms_binned.sum(axis=1), alpha = 0.5,bins = 19) axes3[i].set(yscale='log') dms_binned_norm = np.zeros((19,200)) moments = np.zeros(19) for j in range(19): dms_binned_norm[j],_= np.histogram(m_bin_edges[:-1], m_bin_edges, weights = dms_binned[j], density=True); # print('number of -ve dm in {} for {}: {:,}'.format(t_dict[j],label_range[i], dms_binned[j,:100].sum())) # print('number of +ve dm in {} for {}: {:,}'.format(t_dict[j],label_range[i], dms_binned[j,100:].sum())) moments = calc_moments(m_bin_centres,dms_binned_norm) for idx, ax2 in enumerate(axes2.ravel()): ax2.plot(t_bin_chunk_centres, moments[idx], lw = 0.5, marker = 'o', label = label_range_val[i], color = cmap(i/10.0)); # ax2.legend() ax2.set(xlabel='mjd', ylabel = label_moment[idx]) ax2.axhline(y=0, lw=0.5, ls = '--') # ax2.title.set_text(label_moment[idx]) ax.set(xlabel='mjd', ylabel = 'structure function') if save == True: # fig.savefig('SF_{}.pdf'.format(key),bbox_inches='tight') fig2.savefig('moments_{}.pdf'.format(key),bbox_inches='tight') return fig, ax, fig2, axes2, fig3, axes3 #------------------------------------------------------------------------------ # #------------------------------------------------------------------------------ def plot_sf_ensemble(self, save = False): fig, ax = plt.subplots(1,1,figsize = (16,8)) dms_binned_tot = np.zeros((8,19,200)) for i in range(8): dms_binned, dts_binned, t_bin_edges, t_bin_chunk, t_bin_chunk_centres, m_bin_edges, m_bin_centres, t_dict = self.bin_dtdm(t_max = 3011, t_spacing='log', m_spacing='log', read_in=i+1, key=key) dms_binned_tot[i] = dms_binned dms_binned_tot = dms_binned_tot.sum(axis=0) SF = (((m_bin_centres**2)*dms_binned_tot).sum(axis=1)/dms_binned_tot.sum(axis=1))**0.5 ax.plot(t_bin_chunk_centres,SF, lw = 0.5, marker = 'o') ax.set(yscale='log',xscale='log') ax.set(xlabel='mjd',ylabel = 'structure function') if save == True: fig.savefig('SF_ensemble.pdf',bbox_inches='tight') #------------------------------------------------------------------------------ # #------------------------------------------------------------------------------ def plot_series(df, uids, marker_dict={}, survey_dict={}, plt_color={}, sid=None, bands='r', show_outliers=False, figax=None, **kwargs): """ Plot lightcurve of given objects Parameters ---------- uids : array_like uids of objects to plot sid : int Only plot data from given survey 1 = SSS_r1, 3 = SSS_r2, 5 = SDSS, 7 = PS1, 11 = ZTF """ if np.issubdtype(type(uids),np.integer): uids = [uids] if figax is None: fig, axes = plt.subplots(len(uids),1,figsize = (25,3*len(uids)), sharex=True) else: fig, axes = figax if len(uids)==1: axes=[axes] for uid, ax in zip(uids,axes): single_obj = df.loc[uid].sort_values('mjd') for band in bands: single_band = single_obj[single_obj['band']==band] if sid is not None: # Restrict data to a single survey single_band = single_band[single_band['sid'].isin(sid)] for sid_ in single_band['sid'].unique(): x = single_band[single_band['sid']==sid_] ax.errorbar(x['mjd'], x['mag'], yerr = x['magerr'], lw = 0.5, markersize = 3, marker = marker_dict[sid_], label = survey_dict[sid_]+' '+band, color = plt_color[band]) if show_outliers: outlier_mask = single_obj['outlier'].values & single_obj['sid'].isin(sid) for band in bands: mask = single_obj['band']==band ax.scatter(single_obj['mjd'][outlier_mask & mask], single_obj['mag'][outlier_mask & mask], color = plt_color[band], marker="*", zorder=3, s=200)#, edgecolor='k', linewidths=1) # else: # # Plot a single band # if sid is not None: # # Restrict data to a single survey # single_obj = single_obj[single_obj['sid']==sid] # for sid_ in single_obj['sid'].unique(): # x = single_obj[single_obj['sid']==sid_] # ax.errorbar(x['mjd'], x['mag'], yerr = x['magerr'], lw = 0, elinewidth=0.7, marker = self.marker_dict[sid_], label = self.survey_dict[sid_]+' '+filtercodes, color = self.plt_color[filtercodes]) # mean = single_obj['mag'].mean() # ax.axhline(y=mean, color='k', ls='--', lw=0.4, dashes=(50, 20)) # """ # requires having computed MAD previously # """ # mjd, mag, MAD = single_obj.loc[single_obj['MAD']>0.25, ['mjd','mag', 'MAD']].values.T # ax.scatter(mjd, mag, s=100) # string = ', '.join(['{:.3f}' for _ in MAD]).format(*MAD) # # ax.text(0.02, 0.8, 'MAD max: {:.2f}'.format(np.max(MAD)), transform=ax.transAxes, fontsize=10) # ax.text(0.02, 0.8, string, transform=ax.transAxes, fontsize=10) # mu, std = self.df_grouped.loc[uid,['mag_mean','mag_std']].values.T # axis.axhline((mu-5*std),lw=0.5) # axis.axhline((mu+5*std),lw=0.5) # ax2 = ax.twinx() # ax2.set(ylim=np.array(ax.get_ylim())-mean, ylabel=r'$\mathrm{mag} - \overline{\mathrm{mag}}$') # ax2.invert_yaxis() ax.invert_yaxis() ax.set(xlabel='MJD', ylabel='mag', **kwargs) ax.text(0.02, 0.9, 'uid: {}'.format(uid), transform=ax.transAxes, fontsize=10) plt.subplots_adjust(hspace=0) return fig, axes #------------------------------------------------------------------------------ # #------------------------------------------------------------------------------ def plot_series_bokeh(self, uids, survey=None, filtercodes=None): """ Plot lightcurve of given objects using bokeh Parameters ---------- uids : array_like uids of objects to plot catalogue : int Only plot data from given survey survey : 1 = SDSS, 2 = PS, 3 = ZTF """ plots = [] for uid in uids: single_obj = self.df.loc[uid] if survey is not None: single_obj = single_obj[single_obj['catalogue']==survey] p = figure(title='uid: {}'.format(uid), x_axis_label='mjd', y_axis_label='r mag', plot_width=1000, plot_height=400) for cat in single_obj['catalogue'].unique(): mjd, mag = single_obj[single_obj['catalogue']==cat][['mjd','mag']].sort_values('mjd').values.T p.scatter(x=mjd, y=mag, legend_label=self.survey_dict[cat], marker=self.marker_dict_bokeh[cat], color=self.plt_color_bokeh[filtercodes]) p.line (x=mjd, y=mag, line_width=0.5, color=self.plt_color_bokeh[filtercodes]) p.y_range.flipped = True plots.append(p) show(column(plots)) #------------------------------------------------------------------------------ # #------------------------------------------------------------------------------ def plot_property_distributions(self, prop_range_dict, n_width, n_bins = 250, separate_catalogues = True): """ Parameters ---------- prop_range_dict : dict dictionary of keys and their ranges to be used in histogram n_width : int width of subplot n_bins : int number of bins to use in histogram separate_catalogues : boolean if True, plot data from each survey separately. """ m = -( -len(prop_range_dict) // n_width ) fig, axes = plt.subplots(m, n_width, figsize = (5*n_width,5*m)) cat_label_dict = {1:'SDSS', 2:'PanSTARRS', 3:'ZTF'} for property_name, ax, in zip(prop_range_dict, axes.ravel()): if separate_catalogues == True: for cat, color in zip(self.cat_list,'krb'): self.properties[self.properties['catalogue']==cat][property_name].hist(bins = n_bins, ax=ax, alpha = 0.3, color = color, label = cat_label_dict[cat], range=prop_range_dict[property_name]); ax.legend() elif separate_catalogues == False: self.properties[property_name].hist(bins = 250, ax=ax, alpha = 0.3, range=prop_range_dict[property_name]); else: print('Error, seperate_catalogues must be boolean') ax.set(title = property_name)