def main(night_name=None, fpfile=None, hcfiles=None): # ---------------------------------------------------------------------- # Set up # ---------------------------------------------------------------------- # get parameters from config files/run time args/load paths + calibdb p = spirouStartup.Begin(recipe=__NAME__) if hcfiles is None or fpfile is None: names, types = ['fpfile', 'hcfiles'], [str, str] customargs = spirouStartup.GetCustomFromRuntime(p, [0, 1], types, names, last_multi=True) else: customargs = dict(hcfiles=hcfiles, fpfile=fpfile) # get parameters from configuration files and run time arguments p = spirouStartup.LoadArguments(p, night_name, customargs=customargs, mainfitsdir='reduced', mainfitsfile='hcfiles') # ---------------------------------------------------------------------- # Construct reference filename and get fiber type # ---------------------------------------------------------------------- p, fpfitsfilename = spirouStartup.SingleFileSetup(p, filename=p['FPFILE']) fiber1 = str(p['FIBER']) p, hcfilenames = spirouStartup.MultiFileSetup(p, files=p['HCFILES']) fiber2 = str(p['FIBER']) # set the hcfilename to the first hcfilenames hcfitsfilename = hcfilenames[0] # ---------------------------------------------------------------------- # Once we have checked the e2dsfile we can load calibDB # ---------------------------------------------------------------------- # as we have custom arguments need to load the calibration database p = spirouStartup.LoadCalibDB(p) # ---------------------------------------------------------------------- # Have to check that the fibers match # ---------------------------------------------------------------------- if fiber1 == fiber2: p['FIBER'] = fiber1 fsource = __NAME__ + '/main() & spirouStartup.GetFiberType()' p.set_source('FIBER', fsource) else: emsg = 'Fiber not matching for {0} and {1}, should be the same' eargs = [hcfitsfilename, fpfitsfilename] WLOG(p, 'error', emsg.format(*eargs)) # set the fiber type p['FIB_TYP'] = [p['FIBER']] p.set_source('FIB_TYP', __NAME__ + '/main()') # set find line mode find_lines_mode = p['HC_FIND_LINES_MODE'] # ---------------------------------------------------------------------- # Read image file # ---------------------------------------------------------------------- # read and combine all HC files except the first (fpfitsfilename) rargs = [p, 'add', hcfitsfilename, hcfilenames[1:]] p, hcdata, hchdr = spirouImage.ReadImageAndCombine(*rargs) # read first file (fpfitsfilename) fpdata, fphdr, _, _ = spirouImage.ReadImage(p, fpfitsfilename) # add data and hdr to loc loc = ParamDict() loc['HCDATA'], loc['HCHDR'] = hcdata, hchdr loc['FPDATA'], loc['FPHDR'] = fpdata, fphdr # set the source sources = ['HCDATA', 'HCHDR'] loc.set_sources(sources, 'spirouImage.ReadImageAndCombine()') sources = ['FPDATA', 'FPHDR'] loc.set_sources(sources, 'spirouImage.ReadImage()') # ---------------------------------------------------------------------- # Get basic image properties for reference file # ---------------------------------------------------------------------- # get sig det value p = spirouImage.GetSigdet(p, hchdr, name='sigdet') # get exposure time p = spirouImage.GetExpTime(p, hchdr, name='exptime') # get gain p = spirouImage.GetGain(p, hchdr, name='gain') # get acquisition time p = spirouImage.GetAcqTime(p, hchdr, name='acqtime', kind='julian') bjdref = p['ACQTIME'] # set sigdet and conad keywords (sigdet is changed later) p['KW_CCD_SIGDET'][1] = p['SIGDET'] p['KW_CCD_CONAD'][1] = p['GAIN'] # get lamp parameters p = spirouTHORCA.GetLampParams(p, hchdr) # ---------------------------------------------------------------------- # Obtain the flat # ---------------------------------------------------------------------- # get the flat p, loc = spirouFLAT.GetFlat(p, loc, hchdr) # ---------------------------------------------------------------------- # Read blaze # ---------------------------------------------------------------------- # get tilts p, loc['BLAZE'] = spirouImage.ReadBlazeFile(p, hchdr) loc.set_source('BLAZE', __NAME__ + '/main() + /spirouImage.ReadBlazeFile') # correct the data with the flat # TODO: Should this be used? # log # WLOG(p, '', 'Applying flat correction') # loc['HCDATA'] = loc['HCDATA']/loc['FLAT'] # loc['FPDATA'] = loc['FPDATA']/loc['FLAT'] # ---------------------------------------------------------------------- # Start plotting session # ---------------------------------------------------------------------- if p['DRS_PLOT'] > 0: # start interactive plot sPlt.start_interactive_session(p) # ---------------------------------------------------------------------- # loop around fiber type # ---------------------------------------------------------------------- for fiber in p['FIB_TYP']: # set fiber type for inside loop p['FIBER'] = fiber # ------------------------------------------------------------------ # Instrumental drift computation (if previous solution exists) # ------------------------------------------------------------------ # get key keydb = 'HCREF_{0}'.format(p['FIBER']) # check for key in calibDB if keydb in p['CALIBDB'].keys(): # log process wmsg = ('Doing Instrumental drift computation from previous ' 'calibration') WLOG(p, '', wmsg) # calculate instrument drift loc = spirouTHORCA.CalcInstrumentDrift(p, loc) # ------------------------------------------------------------------ # Wave solution # ------------------------------------------------------------------ # log message for loop wmsg = 'Processing Wavelength Calibration for Fiber {0}' WLOG(p, 'info', wmsg.format(p['FIBER'])) # ------------------------------------------------------------------ # Part 1 of cal_HC # ------------------------------------------------------------------ p, loc = cal_HC_E2DS_spirou.part1(p, loc, mode=find_lines_mode) # ------------------------------------------------------------------ # FP solution # ------------------------------------------------------------------ # log message wmsg = 'Calculating FP wave solution' WLOG(p, '', wmsg) # calculate FP wave solution # spirouTHORCA.FPWaveSolution(p, loc, mode=find_lines_mode) spirouTHORCA.FPWaveSolutionNew(p, loc) # ------------------------------------------------------------------ # FP solution plots # ------------------------------------------------------------------ if p['DRS_PLOT'] > 0: # Plot the FP extracted spectrum against wavelength solution sPlt.wave_plot_final_fp_order(p, loc, iteration=1) # Plot the measured FP cavity width offset against line number sPlt.wave_local_width_offset_plot(p, loc) # Plot the FP line wavelength residuals sPlt.wave_fp_wavelength_residuals(p, loc) # ------------------------------------------------------------------ # Part 2 of cal_HC # ------------------------------------------------------------------ # set params for part2 p['QC_RMS_LITTROW_MAX'] = p['QC_WAVE_RMS_LITTROW_MAX'] p['QC_DEV_LITTROW_MAX'] = p['QC_WAVE_DEV_LITTROW_MAX'] p['IC_HC_N_ORD_START_2'] = min(p['IC_HC_N_ORD_START_2'], p['IC_FP_N_ORD_START']) p['IC_HC_N_ORD_FINAL_2'] = max(p['IC_HC_N_ORD_FINAL_2'], p['IC_FP_N_ORD_FINAL']) # run part 2 # p, loc = part2test(p, loc) p, loc = cal_HC_E2DS_spirou.part2(p, loc) # ---------------------------------------------------------------------- # End plotting session # ---------------------------------------------------------------------- # end interactive session sPlt.end_interactive_session(p) # ---------------------------------------------------------------------- # End Message # ---------------------------------------------------------------------- p = spirouStartup.End(p) # return a copy of locally defined variables in the memory return dict(locals())
def main(night_name=None, files=None): # ---------------------------------------------------------------------- # Set up # ---------------------------------------------------------------------- # get parameters from config files/run time args/load paths + calibdb p = spirouStartup.Begin(recipe=__NAME__) # get parameters from configuration files and run time arguments customargs = spirouStartup.GetCustomFromRuntime(p, [0], [str], ['reffile']) p = spirouStartup.LoadArguments(p, night_name, customargs=customargs, mainfitsfile='reffile', mainfitsdir='reduced') # setup files and get fiber p = spirouStartup.InitialFileSetup(p, calibdb=True) # set the fiber type p['FIB_TYP'] = [p['FIBER']] # ---------------------------------------------------------------------- # Read image file # ---------------------------------------------------------------------- # read the image data gfkwargs = dict(path=p['REDUCED_DIR'], filename=p['REFFILE']) p['REFFILENAME'] = spirouStartup.GetFile(p, **gfkwargs) p.set_source('REFFILENAME', __NAME__ + '/main()') # get the fiber type p['FIBER'] = 'AB' e2ds, hdr, nx, ny = spirouImage.ReadImage(p) # Force A and B to AB solution if p['FIBER'] in ['A', 'B']: wave_fiber = 'AB' else: wave_fiber = p['FIBER'] # get wave image _, wave, _ = spirouImage.GetWaveSolution(p, hdr=hdr, return_wavemap=True, fiber=wave_fiber) blaze = spirouImage.ReadBlazeFile(p) # ---------------------------------------------------------------------- # Get lamp params # ---------------------------------------------------------------------- # get lamp parameters p = spirouTHORCA.GetLampParams(p, hdr) # ---------------------------------------------------------------------- # Get catalogue and fitted line list # ---------------------------------------------------------------------- # load line file (from p['IC_LL_LINE_FILE']) ll_line_cat, ampl_line_cat = spirouImage.ReadLineList(p) # construct fitted lines table filename wavelltbl = spirouConfig.Constants.WAVE_LINE_FILE(p) WLOG(p, '', wavelltbl) # read fitted lines ll_ord, ll_line_fit, ampl_line_fit = np.genfromtxt(wavelltbl, skip_header=4, skip_footer=2, unpack=True, usecols=(0, 1, 3)) # ---------------------------------------------------------------------- # Plots # ---------------------------------------------------------------------- # define line colours col = ['magenta', 'purple'] # get order parity ll_ord_par = np.mod(ll_ord, 2) print(ll_ord_par) col2 = [col[int(x)] for x in ll_ord_par] # start interactive plot sPlt.start_interactive_session(p) plt.figure() for order_num in np.arange(nx): plt.plot(wave[order_num], e2ds[order_num]) # get heights heights = [] for line in range(len(ll_line_cat)): heights.append(200000 + np.max([np.min(e2ds), ampl_line_cat[line]])) # plot ll_line_cat plt.vlines(ll_line_cat, 0, heights, colors='darkgreen', linestyles='dashed') # get heights heights = [] for line in range(len(ll_line_fit)): heights.append(200000 + np.max([np.min(e2ds), ampl_line_fit[line]])) # plot ll_line_fit plt.vlines(ll_line_fit, 0, heights, colors=col2, linestyles='dashdot') plt.xlabel('Wavelength [nm]') plt.ylabel('Flux e-') plt.title(p['REFFILENAME']) # end interactive session # sPlt.end_interactive_session() # old code: # plt.ion() # plt.figure() # # for order_num in np.arange(nx): # plt.plot(wave[order_num], e2ds[order_num]) # # for line in range(len(ll_line_cat)): # plt.vlines(ll_line_cat[line], 0, 200000 + # max(np.min(e2ds), ampl_line_cat[line]), # colors='darkgreen', linestyles='dashed') # # for line in range(len(ll_line_fit)): # plt.vlines(ll_line_fit[line], 0, 200000 + # max(np.min(e2ds), ampl_line_fit[line]), # colors='magenta', linestyles='dashdot') # # plt.xlabel('Wavelength [nm]') # plt.ylabel('Flux e-') # ---------------------------------------------------------------------- # End Message # ---------------------------------------------------------------------- p = spirouStartup.End(p, outputs=None) # return a copy of locally defined variables in the memory return dict(locals())
def main(night_name=None, hcfile=None, fpfiles=None): """ cal_SLIT_spirou.py main function, if night_name and files are None uses arguments from run time i.e.: cal_SLIT_spirou.py [night_directory] [files] :param night_name: string or None, the folder within data raw directory containing files (also reduced directory) i.e. /data/raw/20170710 would be "20170710" but /data/raw/AT5/20180409 would be "AT5/20180409" :param files: string, list or None, the list of files to use for arg_file_names and fitsfilename (if None assumes arg_file_names was set from run time) :return ll: dictionary, containing all the local variables defined in main """ # ---------------------------------------------------------------------- # Set up # ---------------------------------------------------------------------- # get parameters from config files/run time args/load paths + calibdb p = spirouStartup.Begin(recipe=__NAME__) if hcfile is None or fpfiles is None: names, types = ['hcfile', 'fpfiles'], [str, str] customargs = spirouStartup.GetCustomFromRuntime(p, [0, 1], types, names, last_multi=True) else: customargs = dict(hcfile=hcfile, fpfiles=fpfiles) # get parameters from configuration files and run time arguments p = spirouStartup.LoadArguments(p, night_name, customargs=customargs, mainfitsfile='fpfiles') # ---------------------------------------------------------------------- # Construct reference filename and get fiber type # ---------------------------------------------------------------------- p, hcfitsfilename = spirouStartup.SingleFileSetup(p, filename=p['HCFILE']) p, fpfilenames = spirouStartup.MultiFileSetup(p, files=p['FPFILES']) # set fiber (it doesn't matter with the 2D image but we need this to get # the lamp type for FPFILES and HCFILES, AB == C p['FIBER'] = 'AB' p['FIB_TYP'] = [p['FIBER']] fsource = __NAME__ + '/main()' p.set_sources(['FIBER', 'FIB_TYP'], fsource) # set the hcfilename to the first hcfilenames fpfitsfilename = fpfilenames[0] # ---------------------------------------------------------------------- # Once we have checked the e2dsfile we can load calibDB # ---------------------------------------------------------------------- # as we have custom arguments need to load the calibration database p = spirouStartup.LoadCalibDB(p) # add a force plot off p['PLOT_PER_ORDER'] = PLOT_PER_ORDER p.set_source('PLOT_PER_ORDER', __NAME__ + '.main()') # ---------------------------------------------------------------------- # Read FP and HC files # ---------------------------------------------------------------------- # read and combine all FP files except the first (fpfitsfilename) rargs = [p, 'add', fpfitsfilename, fpfilenames[1:]] p, fpdata, fphdr = spirouImage.ReadImageAndCombine(*rargs) # read first file (hcfitsfilename) hcdata, hchdr, _, _ = spirouImage.ReadImage(p, hcfitsfilename) # add data and hdr to loc loc = ParamDict() loc['HCDATA'], loc['HCHDR'] = hcdata, hchdr loc['FPDATA'], loc['FPHDR'] = fpdata, fphdr # set the source sources = ['HCDATA', 'HCHDR'] loc.set_sources(sources, 'spirouImage.ReadImageAndCombine()') sources = ['FPDATA', 'FPHDR'] loc.set_sources(sources, 'spirouImage.ReadImage()') # --------------------------------------------------------------------- # fix for un-preprocessed files # ---------------------------------------------------------------------- hcdata = spirouImage.FixNonPreProcess(p, hcdata) fpdata = spirouImage.FixNonPreProcess(p, fpdata) # ---------------------------------------------------------------------- # Get basic image properties for reference file # ---------------------------------------------------------------------- # get sig det value p = spirouImage.GetSigdet(p, fphdr, name='sigdet') # get exposure time p = spirouImage.GetExpTime(p, fphdr, name='exptime') # get gain p = spirouImage.GetGain(p, fphdr, name='gain') # get lamp parameters p = spirouTHORCA.GetLampParams(p, hchdr) # ---------------------------------------------------------------------- # Correction of DARK # ---------------------------------------------------------------------- # p, hcdatac = spirouImage.CorrectForDark(p, hcdata, hchdr) hcdatac = hcdata p['DARKFILE'] = 'None' # p, fpdatac = spirouImage.CorrectForDark(p, fpdata, fphdr) fpdatac = fpdata # ---------------------------------------------------------------------- # Resize hc data # ---------------------------------------------------------------------- # rotate the image and convert from ADU/s to e- hcdata = spirouImage.ConvertToE(spirouImage.FlipImage(p, hcdatac), p=p) # convert NaN to zeros hcdata0 = np.where(~np.isfinite(hcdata), np.zeros_like(hcdata), hcdata) # resize image bkwargs = dict(xlow=p['IC_CCDX_LOW'], xhigh=p['IC_CCDX_HIGH'], ylow=p['IC_CCDY_LOW'], yhigh=p['IC_CCDY_HIGH'], getshape=False) hcdata2 = spirouImage.ResizeImage(p, hcdata0, **bkwargs) # log change in data size WLOG(p, '', ('HC Image format changed to ' '{0}x{1}').format(*hcdata2.shape)) # ---------------------------------------------------------------------- # Resize fp data # ---------------------------------------------------------------------- # rotate the image and convert from ADU/s to e- fpdata = spirouImage.ConvertToE(spirouImage.FlipImage(p, fpdatac), p=p) # convert NaN to zeros fpdata0 = np.where(~np.isfinite(fpdata), np.zeros_like(fpdata), fpdata) # resize image bkwargs = dict(xlow=p['IC_CCDX_LOW'], xhigh=p['IC_CCDX_HIGH'], ylow=p['IC_CCDY_LOW'], yhigh=p['IC_CCDY_HIGH'], getshape=False) fpdata2 = spirouImage.ResizeImage(p, fpdata0, **bkwargs) # log change in data size WLOG(p, '', ('FP Image format changed to ' '{0}x{1}').format(*fpdata2.shape)) # ---------------------------------------------------------------------- # Correct for the BADPIX mask (set all bad pixels to zero) # ---------------------------------------------------------------------- # p, hcdata2 = spirouImage.CorrectForBadPix(p, hcdata2, hchdr) # p, fpdata2 = spirouImage.CorrectForBadPix(p, fpdata2, fphdr) p['BADPFILE'] = 'None' # save data to loc loc['HCDATA'] = hcdata2 loc.set_source('HCDATA', __NAME__ + '/main()') # save data to loc loc['FPDATA'] = fpdata2 loc.set_source('FPDATA', __NAME__ + '/main()') # ---------------------------------------------------------------------- # Log the number of dead pixels # ---------------------------------------------------------------------- # get the number of bad pixels n_bad_pix = np.nansum(hcdata2 <= 0) n_bad_pix_frac = n_bad_pix * 100 / np.product(hcdata2.shape) # Log number wmsg = 'Nb HC dead pixels = {0} / {1:.2f} %' WLOG(p, 'info', wmsg.format(int(n_bad_pix), n_bad_pix_frac)) # ---------------------------------------------------------------------- # Log the number of dead pixels # ---------------------------------------------------------------------- # get the number of bad pixels n_bad_pix = np.nansum(fpdata2 <= 0) n_bad_pix_frac = n_bad_pix * 100 / np.product(fpdata2.shape) # Log number wmsg = 'Nb FP dead pixels = {0} / {1:.2f} %' WLOG(p, 'info', wmsg.format(int(n_bad_pix), n_bad_pix_frac)) # ------------------------------------------------------------------ # Get localisation coefficients # ------------------------------------------------------------------ # original there is a loop but it is not used --> removed p = spirouImage.FiberParams(p, p['FIBER'], merge=True) # get localisation fit coefficients p, loc = spirouLOCOR.GetCoeffs(p, fphdr, loc) # ------------------------------------------------------------------ # Get master wave solution map # ------------------------------------------------------------------ # get master wave map masterwavefile = spirouDB.GetDatabaseMasterWave(p) # log process wmsg1 = 'Getting master wavelength grid' wmsg2 = '\tFile = {0}'.format(os.path.basename(masterwavefile)) WLOG(p, '', [wmsg1, wmsg2]) # Force A and B to AB solution if p['FIBER'] in ['A', 'B']: wave_fiber = 'AB' else: wave_fiber = p['FIBER'] # read master wave map wout = spirouImage.GetWaveSolution(p, filename=masterwavefile, return_wavemap=True, quiet=True, return_header=True, fiber=wave_fiber) loc['MASTERWAVEP'], loc['MASTERWAVE'] = wout[:2] loc['MASTERWAVEHDR'], loc['WSOURCE'] = wout[2:] # set sources wsource = ['MASTERWAVEP', 'MASTERWAVE', 'MASTERWAVEHDR'] loc.set_sources(wsource, 'spirouImage.GetWaveSolution()') # ---------------------------------------------------------------------- # Read UNe solution # ---------------------------------------------------------------------- wave_u_ne, amp_u_ne = spirouImage.ReadLineList(p) loc['LL_LINE'], loc['AMPL_LINE'] = wave_u_ne, amp_u_ne source = __NAME__ + '.main() + spirouImage.ReadLineList()' loc.set_sources(['LL_LINE', 'AMPL_LINE'], source) # ---------------------------------------------------------------------- # Read cavity length file # ---------------------------------------------------------------------- loc['CAVITY_LEN_COEFFS'] = spirouImage.ReadCavityLength(p) source = __NAME__ + '.main() + spirouImage.ReadCavityLength()' loc.set_source('CAVITY_LEN_COEFFS', source) # ------------------------------------------------------------------ # Calculate shape map # ------------------------------------------------------------------ loc = spirouImage.GetShapeMap(p, loc) # ------------------------------------------------------------------ # Plotting # ------------------------------------------------------------------ if p['DRS_PLOT'] > 0: # plots setup: start interactive plot sPlt.start_interactive_session(p) # plot the shape process for one order sPlt.slit_shape_angle_plot(p, loc) # end interactive section sPlt.end_interactive_session(p) # ---------------------------------------------------------------------- # Quality control # ---------------------------------------------------------------------- # TODO: Decide on some quality control criteria? # set passed variable and fail message list passed, fail_msg = True, [] qc_values, qc_names, qc_logic, qc_pass = [], [], [], [] # finally log the failed messages and set QC = 1 if we pass the # quality control QC = 0 if we fail quality control if passed: WLOG(p, 'info', 'QUALITY CONTROL SUCCESSFUL - Well Done -') p['QC'] = 1 p.set_source('QC', __NAME__ + '/main()') else: for farg in fail_msg: wmsg = 'QUALITY CONTROL FAILED: {0}' WLOG(p, 'warning', wmsg.format(farg)) p['QC'] = 0 p.set_source('QC', __NAME__ + '/main()') # add to qc header lists qc_values.append('None') qc_names.append('None') qc_logic.append('None') qc_pass.append(1) # store in qc_params qc_params = [qc_names, qc_values, qc_logic, qc_pass] # ------------------------------------------------------------------ # Writing DXMAP to file # ------------------------------------------------------------------ # get the raw tilt file name raw_shape_file = os.path.basename(p['FITSFILENAME']) # construct file name and path shapefits, tag = spirouConfig.Constants.SLIT_XSHAPE_FILE(p) shapefitsname = os.path.basename(shapefits) # Log that we are saving tilt file wmsg = 'Saving shape information in file: {0}' WLOG(p, '', wmsg.format(shapefitsname)) # Copy keys from fits file hdict = spirouImage.CopyOriginalKeys(fphdr) # add version number hdict = spirouImage.AddKey(p, hdict, p['KW_VERSION']) hdict = spirouImage.AddKey(p, hdict, p['KW_DRS_DATE'], value=p['DRS_DATE']) hdict = spirouImage.AddKey(p, hdict, p['KW_DATE_NOW'], value=p['DATE_NOW']) hdict = spirouImage.AddKey(p, hdict, p['KW_PID'], value=p['PID']) hdict = spirouImage.AddKey(p, hdict, p['KW_OUTPUT'], value=tag) hdict = spirouImage.AddKey(p, hdict, p['KW_CDBDARK'], value=p['DARKFILE']) hdict = spirouImage.AddKey(p, hdict, p['KW_CDBBAD'], value=p['BADPFILE']) hdict = spirouImage.AddKey(p, hdict, p['KW_CDBLOCO'], value=p['LOCOFILE']) hdict = spirouImage.AddKey1DList(p, hdict, p['KW_INFILE1'], dim1name='hcfile', values=p['HCFILE']) hdict = spirouImage.AddKey1DList(p, hdict, p['KW_INFILE2'], dim1name='fpfile', values=p['FPFILES']) # add qc parameters hdict = spirouImage.AddKey(p, hdict, p['KW_DRS_QC'], value=p['QC']) hdict = spirouImage.AddQCKeys(p, hdict, qc_params) # write tilt file to file p = spirouImage.WriteImage(p, shapefits, loc['DXMAP'], hdict) # ------------------------------------------------------------------ # Writing sanity check files # ------------------------------------------------------------------ if p['SHAPE_DEBUG_OUTPUTS']: # log WLOG(p, '', 'Saving debug sanity check files') # construct file names input_fp_file, tag1 = spirouConfig.Constants.SLIT_SHAPE_IN_FP_FILE(p) output_fp_file, tag2 = spirouConfig.Constants.SLIT_SHAPE_OUT_FP_FILE(p) input_hc_file, tag3 = spirouConfig.Constants.SLIT_SHAPE_IN_HC_FILE(p) output_hc_file, tag4 = spirouConfig.Constants.SLIT_SHAPE_OUT_HC_FILE(p) overlap_file, tag5 = spirouConfig.Constants.SLIT_SHAPE_OVERLAP_FILE(p) # write input fp file hdict = spirouImage.AddKey(p, hdict, p['KW_OUTPUT'], value=tag1) p = spirouImage.WriteImage(p, input_fp_file, loc['FPDATA'], hdict) # write output fp file hdict = spirouImage.AddKey(p, hdict, p['KW_OUTPUT'], value=tag2) p = spirouImage.WriteImage(p, output_fp_file, loc['FPDATA2'], hdict) # write input fp file hdict = spirouImage.AddKey(p, hdict, p['KW_OUTPUT'], value=tag3) p = spirouImage.WriteImage(p, input_hc_file, loc['HCDATA'], hdict) # write output fp file hdict = spirouImage.AddKey(p, hdict, p['KW_OUTPUT'], value=tag4) p = spirouImage.WriteImage(p, output_hc_file, loc['HCDATA2'], hdict) # write overlap file hdict = spirouImage.AddKey(p, hdict, p['KW_OUTPUT'], value=tag5) p = spirouImage.WriteImage(p, overlap_file, loc['ORDER_OVERLAP'], hdict) # ---------------------------------------------------------------------- # Move to calibDB and update calibDB # ---------------------------------------------------------------------- if p['QC']: keydb = 'SHAPE' # copy shape file to the calibDB folder spirouDB.PutCalibFile(p, shapefits) # update the master calib DB file with new key spirouDB.UpdateCalibMaster(p, keydb, shapefitsname, fphdr) # ---------------------------------------------------------------------- # End Message # ---------------------------------------------------------------------- p = spirouStartup.End(p) # return a copy of locally defined variables in the memory return dict(locals())
def main(night_name=None, files=None): """ cal_HC_E2DS.py main function, if night_name and files are None uses arguments from run time i.e.: cal_DARK_spirou.py [night_directory] [fitsfilename] :param night_name: string or None, the folder within data raw directory containing files (also reduced directory) i.e. /data/raw/20170710 would be "20170710" but /data/raw/AT5/20180409 would be "AT5/20180409" :param files: string, list or None, the list of files to use for arg_file_names and fitsfilename (if None assumes arg_file_names was set from run time) :return ll: dictionary, containing all the local variables defined in main """ # ---------------------------------------------------------------------- # Set up # ---------------------------------------------------------------------- # get parameters from config files/run time args/load paths + calibdb p = spirouStartup.Begin(recipe=__NAME__) # get parameters from configuration files and run time arguments p = spirouStartup.LoadArguments(p, night_name, files, mainfitsdir='reduced') # setup files and get fiber p = spirouStartup.InitialFileSetup(p, calibdb=True) # set the fiber type p['FIB_TYP'] = [p['FIBER']] p.set_source('FIB_TYP', __NAME__ + '/main()') # ---------------------------------------------------------------------- # Read image file # ---------------------------------------------------------------------- # read and combine all files p, hcdata, hchdr = spirouImage.ReadImageAndCombine(p, 'add') # add data and hdr to loc loc = ParamDict() loc['HCDATA'], loc['HCHDR'] = hcdata, hchdr # set the source sources = ['HCDATA', 'HCHDR'] loc.set_sources(sources, 'spirouImage.ReadImageAndCombine()') # ---------------------------------------------------------------------- # Get basic parameters # ---------------------------------------------------------------------- # get sig det value p = spirouImage.GetSigdet(p, loc['HCHDR'], name='sigdet') # get exposure time p = spirouImage.GetExpTime(p, loc['HCHDR'], name='exptime') # get gain p = spirouImage.GetGain(p, loc['HCHDR'], name='gain') # get acquisition time p = spirouImage.GetAcqTime(p, loc['HCHDR'], name='ACQTIME', kind='julian') bjdref = p['ACQTIME'] # set sigdet and conad keywords (sigdet is changed later) p['KW_CCD_SIGDET'][1] = p['SIGDET'] p['KW_CCD_CONAD'][1] = p['GAIN'] # get lamp parameters p = spirouTHORCA.GetLampParams(p, loc['HCHDR']) # get number of orders # we always get fibre A number because AB is doubled in constants file loc['NBO'] = p['QC_LOC_NBO_FPALL']['A'] loc.set_source('NBO', __NAME__ + '.main()') # get number of pixels in x from hcdata size loc['NBPIX'] = loc['HCDATA'].shape[1] loc.set_source('NBPIX', __NAME__ + '.main()') # ---------------------------------------------------------------------- # Read blaze # ---------------------------------------------------------------------- # get tilts loc['BLAZE'] = spirouImage.ReadBlazeFile(p, hchdr) loc.set_source('BLAZE', __NAME__ + '/main() + /spirouImage.ReadBlazeFile') # ---------------------------------------------------------------------- # Read wave solution # ---------------------------------------------------------------------- # wavelength file; we will use the polynomial terms in its header, # NOT the pixel values that would need to be interpolated # getting header info with wavelength polynomials # set source of wave file wsource = __NAME__ + '/main() + /spirouImage.GetWaveSolution' # Force A and B to AB solution if p['FIBER'] in ['A', 'B']: wave_fiber = 'AB' else: wave_fiber = p['FIBER'] # get wave image wout = spirouImage.GetWaveSolution(p, hdr=hchdr, return_wavemap=True, return_filename=True, fiber=wave_fiber) loc['WAVEPARAMS'], loc['WAVE_INIT'], loc['WAVEFILE'], loc['WSOURCE'] = wout loc.set_sources(['WAVE_INIT', 'WAVEFILE', 'WAVEPARAMS', 'WSOURCE'], wsource) # ---------------------------------------------------------------------- # Check that wave parameters are consistent with "ic_ll_degr_fit" # ---------------------------------------------------------------------- loc = spirouImage.CheckWaveSolConsistency(p, loc) # ---------------------------------------------------------------------- # Read UNe solution # ---------------------------------------------------------------------- wave_u_ne, amp_u_ne = spirouImage.ReadLineList(p) loc['LL_LINE'], loc['AMPL_LINE'] = wave_u_ne, amp_u_ne source = __NAME__ + '.main() + spirouImage.ReadLineList()' loc.set_sources(['ll_line', 'ampl_line'], source) # ---------------------------------------------------------------------- # Generate wave map from wave solution # ---------------------------------------------------------------------- loc = spirouWAVE.generate_wave_map(p, loc) # ---------------------------------------------------------------------- # Find Gaussian Peaks in HC spectrum # ---------------------------------------------------------------------- loc = spirouWAVE.find_hc_gauss_peaks(p, loc) # ---------------------------------------------------------------------- # Start plotting session # ---------------------------------------------------------------------- if p['DRS_PLOT'] > 0: # start interactive plot sPlt.start_interactive_session(p) # ---------------------------------------------------------------------- # Fit Gaussian peaks (in triplets) to # ---------------------------------------------------------------------- loc = spirouWAVE.fit_gaussian_triplets(p, loc) # ---------------------------------------------------------------------- # Generate Resolution map and line profiles # ---------------------------------------------------------------------- # log progress wmsg = 'Generating resolution map and ' # generate resolution map loc = spirouWAVE.generate_resolution_map(p, loc) # map line profile map if p['DRS_PLOT'] > 0: sPlt.wave_ea_plot_line_profiles(p, loc) # ---------------------------------------------------------------------- # End plotting session # ---------------------------------------------------------------------- # end interactive session if p['DRS_PLOT'] > 0: sPlt.end_interactive_session(p) # ---------------------------------------------------------------------- # Quality control # ---------------------------------------------------------------------- passed, fail_msg = True, [] qc_values, qc_names, qc_logic, qc_pass = [], [], [], [] # quality control on sigma clip (sig1 > qc_hc_wave_sigma_max if loc['SIG1'] > p['QC_HC_WAVE_SIGMA_MAX']: fmsg = 'Sigma too high ({0:.5f} > {1:.5f})' fail_msg.append(fmsg.format(loc['SIG1'], p['QC_HC_WAVE_SIGMA_MAX'])) passed = False qc_pass.append(0) else: qc_pass.append(1) # add to qc header lists qc_values.append(loc['SIG1']) qc_names.append('SIG1') qc_logic.append('SIG1 > {0:.2f}'.format(p['QC_HC_WAVE_SIGMA_MAX'])) # ---------------------------------------------------------------------- # check the difference between consecutive orders is always positive # get the differences wave_diff = loc['WAVE_MAP2'][1:]-loc['WAVE_MAP2'][:-1] if np.min(wave_diff) < 0: fmsg = 'Negative wavelength difference between orders' fail_msg.append(fmsg) passed = False qc_pass.append(0) else: qc_pass.append(1) # add to qc header lists qc_values.append(np.min(wave_diff)) qc_names.append('MIN WAVE DIFF') qc_logic.append('MIN WAVE DIFF < 0') # ---------------------------------------------------------------------- # finally log the failed messages and set QC = 1 if we pass the # quality control QC = 0 if we fail quality control if passed: WLOG(p, 'info', 'QUALITY CONTROL SUCCESSFUL - Well Done -') p['QC'] = 1 p.set_source('QC', __NAME__ + '/main()') else: for farg in fail_msg: wmsg = 'QUALITY CONTROL FAILED: {0}' WLOG(p, 'warning', wmsg.format(farg)) p['QC'] = 0 p.set_source('QC', __NAME__ + '/main()') # store in qc_params qc_params = [qc_names, qc_values, qc_logic, qc_pass] # ---------------------------------------------------------------------- # log the global stats # ---------------------------------------------------------------------- # calculate catalog-fit residuals in km/s res_hc =[] sumres_hc = 0.0 sumres2_hc = 0.0 for order in range(loc['NBO']): # get HC line wavelengths for the order order_mask = loc['ORD_T'] == order hc_x_ord = loc['XGAU_T'][order_mask] hc_ll_ord = np.polyval(loc['POLY_WAVE_SOL'][order][::-1],hc_x_ord) hc_ll_cat = loc['WAVE_CATALOG'][order_mask] hc_ll_diff = hc_ll_ord - hc_ll_cat res_hc.append(hc_ll_diff*speed_of_light/hc_ll_cat) sumres_hc += np.nansum(res_hc[order]) sumres2_hc += np.nansum(res_hc[order] ** 2) total_lines_hc = len(np.concatenate(res_hc)) final_mean_hc = sumres_hc/total_lines_hc final_var_hc = (sumres2_hc/total_lines_hc) - (final_mean_hc ** 2) wmsg1 = 'On fiber {0} HC fit line statistic:'.format(p['FIBER']) wargs2 = [final_mean_hc * 1000.0, np.sqrt(final_var_hc) * 1000.0, total_lines_hc, 1000.0 * np.sqrt(final_var_hc / total_lines_hc)] wmsg2 = ('\tmean={0:.3f}[m/s] rms={1:.1f} {2} HC lines (error on mean ' 'value:{3:.4f}[m/s])'.format(*wargs2)) WLOG(p, 'info', [wmsg1, wmsg2]) # ---------------------------------------------------------------------- # Save wave map to file # ---------------------------------------------------------------------- # get base input filenames bfilenames = [] for raw_file in p['ARG_FILE_NAMES']: bfilenames.append(os.path.basename(raw_file)) # get wave filename wavefits, tag1 = spirouConfig.Constants.WAVE_FILE_EA(p) wavefitsname = os.path.basename(wavefits) # log progress WLOG(p, '', 'Saving wave map to {0}'.format(wavefitsname)) # log progress wargs = [p['FIBER'], wavefitsname] wmsg = 'Write wavelength solution for Fiber {0} in {1}' WLOG(p, '', wmsg.format(*wargs)) # write solution to fitsfilename header # copy original keys hdict = spirouImage.CopyOriginalKeys(loc['HCHDR']) # set the version hdict = spirouImage.AddKey(p, hdict, p['KW_VERSION']) hdict = spirouImage.AddKey(p, hdict, p['KW_DRS_DATE'], value=p['DRS_DATE']) hdict = spirouImage.AddKey(p, hdict, p['KW_DATE_NOW'], value=p['DATE_NOW']) hdict = spirouImage.AddKey(p, hdict, p['KW_PID'], value=p['PID']) hdict = spirouImage.AddKey(p, hdict, p['KW_OUTPUT'], value=tag1) # set the input files hdict = spirouImage.AddKey(p, hdict, p['KW_CDBBLAZE'], value=p['BLAZFILE']) # add qc parameters hdict = spirouImage.AddKey(p, hdict, p['KW_DRS_QC'], value=p['QC']) hdict = spirouImage.AddQCKeys(p, hdict, qc_params) # add wave solution date hdict = spirouImage.AddKey(p, hdict, p['KW_WAVE_TIME1'], value=p['MAX_TIME_HUMAN']) hdict = spirouImage.AddKey(p, hdict, p['KW_WAVE_TIME2'], value=p['MAX_TIME_UNIX']) hdict = spirouImage.AddKey(p, hdict, p['KW_WAVE_CODE'], value=__NAME__) hdict = spirouImage.AddKey(p, hdict, p['KW_CDBWAVE'], value=loc['WAVEFILE']) hdict = spirouImage.AddKey(p, hdict, p['KW_WAVESOURCE'], value=loc['WSOURCE']) hdict = spirouImage.AddKey1DList(p, hdict, p['KW_INFILE1'], dim1name='file', values=p['ARG_FILE_NAMES']) # add number of orders hdict = spirouImage.AddKey(p, hdict, p['KW_WAVE_ORD_N'], value=loc['POLY_WAVE_SOL'].shape[0]) # add degree of fit hdict = spirouImage.AddKey(p, hdict, p['KW_WAVE_LL_DEG'], value=loc['POLY_WAVE_SOL'].shape[1]-1) # add wave solution hdict = spirouImage.AddKey2DList(p, hdict, p['KW_WAVE_PARAM'], values=loc['POLY_WAVE_SOL']) # write the wave "spectrum" p = spirouImage.WriteImage(p, wavefits, loc['WAVE_MAP2'], hdict) # get filename for E2DS calibDB copy of FITSFILENAME e2dscopy_filename, tag2 = spirouConfig.Constants.WAVE_E2DS_COPY(p) wargs = [p['FIBER'], os.path.split(e2dscopy_filename)[-1]] wmsg = 'Write reference E2DS spectra for Fiber {0} in {1}' WLOG(p, '', wmsg.format(*wargs)) # make a copy of the E2DS file for the calibBD hdict = spirouImage.AddKey(p, hdict, p['KW_OUTPUT'], value=tag2) p = spirouImage.WriteImage(p, e2dscopy_filename, loc['HCDATA'], hdict) # ---------------------------------------------------------------------- # Save resolution and line profiles to file # ---------------------------------------------------------------------- raw_infile = os.path.basename(p['FITSFILENAME']) # get wave filename resfits, tag3 = spirouConfig.Constants.WAVE_RES_FILE_EA(p) resfitsname = os.path.basename(resfits) WLOG(p, '', 'Saving wave resmap to {0}'.format(resfitsname)) # make a copy of the E2DS file for the calibBD # set the version hdict = spirouImage.AddKey(p, hdict, p['KW_VERSION']) hdict = spirouImage.AddKey(p, hdict, p['KW_DRS_DATE'], value=p['DRS_DATE']) hdict = spirouImage.AddKey(p, hdict, p['KW_DATE_NOW'], value=p['DATE_NOW']) hdict = spirouImage.AddKey(p, hdict, p['KW_OUTPUT'], value=tag3) # get res data in correct format resdata, hdicts = spirouTHORCA.GenerateResFiles(p, loc, hdict) # save to file p = spirouImage.WriteImageMulti(p, resfits, resdata, hdicts=hdicts) # ---------------------------------------------------------------------- # Update calibDB # ---------------------------------------------------------------------- if p['QC']: # set the wave key keydb = 'WAVE_{0}'.format(p['FIBER']) # copy wave file to calibDB folder spirouDB.PutCalibFile(p, wavefits) # update the master calib DB file with new key spirouDB.UpdateCalibMaster(p, keydb, wavefitsname, loc['HCHDR']) # set the hcref key keydb = 'HCREF_{0}'.format(p['FIBER']) # copy wave file to calibDB folder spirouDB.PutCalibFile(p, e2dscopy_filename) # update the master calib DB file with new key e2dscopyfits = os.path.split(e2dscopy_filename)[-1] spirouDB.UpdateCalibMaster(p, keydb, e2dscopyfits, loc['HCHDR']) # ---------------------------------------------------------------------- # Update header of current files # ---------------------------------------------------------------------- # only copy over if QC passed if p['QC']: rdir = os.path.dirname(wavefits) # loop around hc files and update header with for rawhcfile in p['ARG_FILE_NAMES']: hcfile = os.path.join(rdir, rawhcfile) raw_infilepath1 = os.path.join(p['ARG_FILE_DIR'], hcfile) p = spirouImage.UpdateWaveSolutionHC(p, loc, raw_infilepath1) # ---------------------------------------------------------------------- # End Message # ---------------------------------------------------------------------- p = spirouStartup.End(p) # return a copy of locally defined variables in the memory return dict(locals())
def main(night_name=None, hcfile=None, fpfiles=None): """ cal_SLIT_spirou.py main function, if night_name and files are None uses arguments from run time i.e.: cal_SLIT_spirou.py [night_directory] [files] :param night_name: string or None, the folder within data raw directory containing files (also reduced directory) i.e. /data/raw/20170710 would be "20170710" but /data/raw/AT5/20180409 would be "AT5/20180409" :param files: string, list or None, the list of files to use for arg_file_names and fitsfilename (if None assumes arg_file_names was set from run time) :return ll: dictionary, containing all the local variables defined in main """ # ---------------------------------------------------------------------- # Set up # ---------------------------------------------------------------------- # get parameters from config files/run time args/load paths + calibdb p = spirouStartup.Begin(recipe=__NAME__) if hcfile is None or fpfiles is None: names, types = ['hcfile', 'fpfiles'], [str, str] customargs = spirouStartup.GetCustomFromRuntime(p, [0, 1], types, names, last_multi=True) else: customargs = dict(hcfile=hcfile, fpfile=fpfiles) # get parameters from configuration files and run time arguments p = spirouStartup.LoadArguments(p, night_name, customargs=customargs, mainfitsfile='fpfiles') # ---------------------------------------------------------------------- # Construct reference filename and get fiber type # ---------------------------------------------------------------------- p, hcfitsfilename = spirouStartup.SingleFileSetup(p, filename=p['HCFILE']) p, fpfitsfiles = spirouStartup.MultiFileSetup(p, files=p['FPFILES']) # set fiber (it doesn't matter with the 2D image but we need this to get # the lamp type for FPFILES and HCFILES, AB == C p['FIBER'] = 'AB' p['FIB_TYP'] = [p['FIBER']] fsource = __NAME__ + '/main()' p.set_sources(['FIBER', 'FIB_TYP'], fsource) # ---------------------------------------------------------------------- # Once we have checked the e2dsfile we can load calibDB # ---------------------------------------------------------------------- # as we have custom arguments need to load the calibration database p = spirouStartup.LoadCalibDB(p) # add a force plot off p['PLOT_PER_ORDER'] = PLOT_PER_ORDER p.set_source('PLOT_PER_ORDER', __NAME__ + '.main()') # ---------------------------------------------------------------------- # Read FP and HC files # ---------------------------------------------------------------------- # read input fp and hc data rkwargs = dict(filename=fpfitsfiles[0], filenames=fpfitsfiles[1:], framemath='add') p, fpdata, fphdr = spirouImage.ReadImageAndCombine(p, **rkwargs) hcdata, hchdr, _, _ = spirouImage.ReadImage(p, hcfitsfilename) # add data and hdr to loc loc = ParamDict() loc['HCDATA'], loc['HCHDR'] = hcdata, hchdr loc['FPDATA'], loc['FPHDR'] = fpdata, fphdr # set the source sources = ['HCDATA', 'HCHDR'] loc.set_sources(sources, 'spirouImage.ReadImage()') sources = ['FPDATA', 'FPHDR'] loc.set_sources(sources, 'spirouImage.ReadImage()') # --------------------------------------------------------------------- # fix for un-preprocessed files # ---------------------------------------------------------------------- hcdata = spirouImage.FixNonPreProcess(p, hcdata) fpdata = spirouImage.FixNonPreProcess(p, fpdata) # ---------------------------------------------------------------------- # Once we have checked the e2dsfile we can load calibDB # ---------------------------------------------------------------------- # as we have custom arguments need to load the calibration database p = spirouStartup.LoadCalibDB(p) # add a force plot off p['PLOT_PER_ORDER'] = PLOT_PER_ORDER p.set_source('PLOT_PER_ORDER', __NAME__ + '.main()') # ---------------------------------------------------------------------- # Get basic image properties for reference file # ---------------------------------------------------------------------- # get sig det value p = spirouImage.GetSigdet(p, fphdr, name='sigdet') # get exposure time p = spirouImage.GetExpTime(p, fphdr, name='exptime') # get gain p = spirouImage.GetGain(p, fphdr, name='gain') # get lamp parameters p = spirouTHORCA.GetLampParams(p, hchdr) # get FP_FP DPRTYPE p = spirouImage.ReadParam(p, fphdr, 'KW_DPRTYPE', 'DPRTYPE', dtype=str) # ---------------------------------------------------------------------- # Correction of reference FP # ---------------------------------------------------------------------- # set the number of frames p['NBFRAMES'] = len(fpfitsfiles) p.set_source('NBFRAMES', __NAME__ + '.main()') # Correction of DARK p, fpdatac = spirouImage.CorrectForDark(p, fpdata, fphdr) # Resize hc data # rotate the image and convert from ADU/s to e- fpdata = spirouImage.ConvertToE(spirouImage.FlipImage(p, fpdatac), p=p) # resize image bkwargs = dict(xlow=p['IC_CCDX_LOW'], xhigh=p['IC_CCDX_HIGH'], ylow=p['IC_CCDY_LOW'], yhigh=p['IC_CCDY_HIGH'], getshape=False) fpdata1 = spirouImage.ResizeImage(p, fpdata, **bkwargs) # log change in data size WLOG(p, '', ('FPref Image format changed to {0}x{1}').format(*fpdata1.shape)) # Correct for the BADPIX mask (set all bad pixels to zero) bargs = [p, fpdata1, fphdr] p, fpdata1 = spirouImage.CorrectForBadPix(*bargs) p, badpixmask = spirouImage.CorrectForBadPix(*bargs, return_map=True) # log progress WLOG(p, '', 'Cleaning FPref hot pixels') # correct hot pixels fpdata1 = spirouEXTOR.CleanHotpix(fpdata1, badpixmask) # add to loc loc['FPDATA1'] = fpdata1 loc.set_source('FPDATA1', __NAME__ + '.main()') # Log the number of dead pixels # get the number of bad pixels with warnings.catch_warnings(record=True) as _: n_bad_pix = np.nansum(fpdata1 <= 0) n_bad_pix_frac = n_bad_pix * 100 / np.product(fpdata1.shape) # Log number wmsg = 'Nb FPref dead pixels = {0} / {1:.2f} %' WLOG(p, 'info', wmsg.format(int(n_bad_pix), n_bad_pix_frac)) # ---------------------------------------------------------------------- # Correction of HC # ---------------------------------------------------------------------- # set the number of frames p['NBFRAMES'] = 1 p.set_source('NBFRAMES', __NAME__ + '.main()') # Correction of DARK p, hcdatac = spirouImage.CorrectForDark(p, hcdata, hchdr) # Resize hc data # rotate the image and convert from ADU/s to e- hcdata = spirouImage.ConvertToE(spirouImage.FlipImage(p, hcdatac), p=p) # resize image bkwargs = dict(xlow=p['IC_CCDX_LOW'], xhigh=p['IC_CCDX_HIGH'], ylow=p['IC_CCDY_LOW'], yhigh=p['IC_CCDY_HIGH'], getshape=False) hcdata1 = spirouImage.ResizeImage(p, hcdata, **bkwargs) # log change in data size WLOG(p, '', ('HC Image format changed to {0}x{1}').format(*hcdata1.shape)) # Correct for the BADPIX mask (set all bad pixels to zero) bargs = [p, hcdata1, hchdr] p, hcdata1 = spirouImage.CorrectForBadPix(*bargs) p, badpixmask = spirouImage.CorrectForBadPix(*bargs, return_map=True) # log progress WLOG(p, '', 'Cleaning HC hot pixels') # correct hot pixels hcdata1 = spirouEXTOR.CleanHotpix(hcdata1, badpixmask) # add to loc loc['HCDATA1'] = hcdata1 loc.set_source('HCDATA1', __NAME__ + '.main()') # Log the number of dead pixels # get the number of bad pixels with warnings.catch_warnings(record=True) as _: n_bad_pix = np.nansum(hcdata1 <= 0) n_bad_pix_frac = n_bad_pix * 100 / np.product(hcdata1.shape) # Log number wmsg = 'Nb HC dead pixels = {0} / {1:.2f} %' WLOG(p, 'info', wmsg.format(int(n_bad_pix), n_bad_pix_frac)) # ------------------------------------------------------------------------- # get all FP_FP files # ------------------------------------------------------------------------- fpfilenames = spirouImage.FindFiles(p, filetype=p['DPRTYPE'], allowedtypes=p['ALLOWED_FP_TYPES']) # convert filenames to a numpy array fpfilenames = np.array(fpfilenames) # julian date to know which file we need to # process together fp_time = np.zeros(len(fpfilenames)) basenames, fp_exp, fp_pp_version, nightnames = [], [], [], [] # log progress WLOG(p, '', 'Reading all fp file headers') # looping through the file headers for it in range(len(fpfilenames)): # log progress wmsg = '\tReading file {0} / {1}' WLOG(p, 'info', wmsg.format(it + 1, len(fpfilenames))) # get fp filename fpfilename = fpfilenames[it] # get night name night_name = os.path.dirname(fpfilenames[it]).split(p['TMP_DIR'])[-1] # read data data_it, hdr_it, _, _ = spirouImage.ReadImage(p, fpfilename) # get header hdr = spirouImage.ReadHeader(p, filepath=fpfilenames[it]) # add MJDATE to dark times fp_time[it] = float(hdr[p['KW_ACQTIME'][0]]) # add other keys (for tabular output) basenames.append(os.path.basename(fpfilenames[it])) nightnames.append(night_name) fp_exp.append(float(hdr[p['KW_EXPTIME'][0]])) fp_pp_version.append(hdr[p['KW_PPVERSION'][0]]) # ------------------------------------------------------------------------- # match files by date # ------------------------------------------------------------------------- # log progress wmsg = 'Matching FP files by observation time (+/- {0} hrs)' WLOG(p, '', wmsg.format(p['DARK_MASTER_MATCH_TIME'])) # get the time threshold time_thres = p['FP_MASTER_MATCH_TIME'] # get items grouped by time matched_id = spirouImage.GroupFilesByTime(p, fp_time, time_thres) # ------------------------------------------------------------------------- # construct the master fp file (+ correct for dark/badpix) # ------------------------------------------------------------------------- cargs = [fpdata1, fpfilenames, matched_id] fpcube, transforms = spirouImage.ConstructMasterFP(p, *cargs) # log process wmsg1 = 'Master FP construction complete.' wmsg2 = '\tAdding {0} group images to form FP master image' WLOG(p, 'info', [wmsg1, wmsg2.format(len(fpcube))]) # sum the cube to make fp data masterfp = np.sum(fpcube, axis=0) # add to loc loc['MASTERFP'] = masterfp loc.set_source('MASTERFP', __NAME__ + '.main()') # ------------------------------------------------------------------ # Get localisation coefficients # ------------------------------------------------------------------ # original there is a loop but it is not used --> removed p = spirouImage.FiberParams(p, p['FIBER'], merge=True) # get localisation fit coefficients p, loc = spirouLOCOR.GetCoeffs(p, fphdr, loc) # ------------------------------------------------------------------ # Get master wave solution map # ------------------------------------------------------------------ # get master wave map masterwavefile = spirouDB.GetDatabaseMasterWave(p) # log process wmsg1 = 'Getting master wavelength grid' wmsg2 = '\tFile = {0}'.format(os.path.basename(masterwavefile)) WLOG(p, '', [wmsg1, wmsg2]) # Force A and B to AB solution if p['FIBER'] in ['A', 'B']: wave_fiber = 'AB' else: wave_fiber = p['FIBER'] # read master wave map wout = spirouImage.GetWaveSolution(p, filename=masterwavefile, return_wavemap=True, quiet=True, return_header=True, fiber=wave_fiber) loc['MASTERWAVEP'], loc['MASTERWAVE'] = wout[:2] loc['MASTERWAVEHDR'], loc['WSOURCE'] = wout[2:] # set sources wsource = ['MASTERWAVEP', 'MASTERWAVE', 'MASTERWAVEHDR'] loc.set_sources(wsource, 'spirouImage.GetWaveSolution()') # ---------------------------------------------------------------------- # Read UNe solution # ---------------------------------------------------------------------- wave_u_ne, amp_u_ne = spirouImage.ReadLineList(p) loc['LL_LINE'], loc['AMPL_LINE'] = wave_u_ne, amp_u_ne source = __NAME__ + '.main() + spirouImage.ReadLineList()' loc.set_sources(['LL_LINE', 'AMPL_LINE'], source) # ---------------------------------------------------------------------- # Read cavity length file # ---------------------------------------------------------------------- loc['CAVITY_LEN_COEFFS'] = spirouImage.ReadCavityLength(p) source = __NAME__ + '.main() + spirouImage.ReadCavityLength()' loc.set_source('CAVITY_LEN_COEFFS', source) # ---------------------------------------------------------------------- # Calculate shape map # ---------------------------------------------------------------------- # calculate dx map loc = spirouImage.GetXShapeMap(p, loc) # if dx map is None we shouldn't continue if loc['DXMAP'] is None: fargs = [ loc['MAXDXMAPINFO'][0], loc['MAXDXMAPINFO'][1], loc['MAXDXMAPSTD'], p['SHAPE_QC_DXMAP_STD'] ] fmsg = ('The std of the dxmap for order {0} y-pixel {1} is too large.' ' std = {2} (limit = {3})'.format(*fargs)) wmsg = 'QUALITY CONTROL FAILED: {0}' WLOG(p, 'warning', wmsg.format(fmsg)) WLOG(p, 'warning', 'Cannot continue. Exiting.') # End Message p = spirouStartup.End(p) # return a copy of locally defined variables in the memory return dict(locals()) # calculate dymap loc = spirouImage.GetYShapeMap(p, loc, fphdr) # ------------------------------------------------------------------ # Need to straighten the dxmap # ------------------------------------------------------------------ # copy it first loc['DXMAP0'] = np.array(loc['DXMAP']) # straighten it loc['DXMAP'] = spirouImage.EATransform(loc['DXMAP'], dymap=loc['DYMAP']) # ------------------------------------------------------------------ # Need to straighten the hc data and fp data for debug # ------------------------------------------------------------------ # log progress WLOG(p, '', 'Shape finding complete. Applying transforms.') # apply very last update of the debananafication tkwargs = dict(dxmap=loc['DXMAP'], dymap=loc['DYMAP']) loc['HCDATA2'] = spirouImage.EATransform(loc['HCDATA1'], **tkwargs) loc['FPDATA2'] = spirouImage.EATransform(loc['FPDATA1'], **tkwargs) loc.set_sources(['HCDATA2', 'FPDATA2'], __NAME__ + '.main()') # ------------------------------------------------------------------ # Plotting # ------------------------------------------------------------------ if p['DRS_PLOT'] > 0: # plots setup: start interactive plot sPlt.start_interactive_session(p) # plot the shape process for one order sPlt.slit_shape_angle_plot(p, loc) # end interactive section sPlt.end_interactive_session(p) # ---------------------------------------------------------------------- # Quality control # ---------------------------------------------------------------------- # TODO: Decide on some quality control criteria? # set passed variable and fail message list passed, fail_msg = True, [] qc_values, qc_names, qc_logic, qc_pass = [], [], [], [] # finally log the failed messages and set QC = 1 if we pass the # quality control QC = 0 if we fail quality control if passed: WLOG(p, 'info', 'QUALITY CONTROL SUCCESSFUL - Well Done -') p['QC'] = 1 p.set_source('QC', __NAME__ + '/main()') else: for farg in fail_msg: wmsg = 'QUALITY CONTROL FAILED: {0}' WLOG(p, 'warning', wmsg.format(farg)) p['QC'] = 0 p.set_source('QC', __NAME__ + '/main()') # add to qc header lists qc_values.append(loc['MAXDXMAPSTD']) qc_names.append('DXMAP STD') qc_logic.append('DXMAP STD < {0}'.format(p['SHAPE_QC_DXMAP_STD'])) qc_pass.append(1) # store in qc_params qc_params = [qc_names, qc_values, qc_logic, qc_pass] # ------------------------------------------------------------------ # Writing FP big table # ------------------------------------------------------------------ # construct big fp table colnames = [ 'FILENAME', 'NIGHT', 'MJDATE', 'EXPTIME', 'PVERSION', 'GROUPID', 'DXREF', 'DYREF', 'A', 'B', 'C', 'D' ] values = [ basenames, nightnames, fp_time, fp_exp, fp_pp_version, matched_id, transforms[:, 0], transforms[:, 1], transforms[:, 2], transforms[:, 3], transforms[:, 4], transforms[:, 5] ] fptable = spirouImage.MakeTable(p, colnames, values) # ------------------------------------------------------------------ # Writing DXMAP to file # ------------------------------------------------------------------ # get the raw tilt file name raw_shape_file = os.path.basename(p['FITSFILENAME']) # construct file name and path shapexfits, tag = spirouConfig.Constants.SLIT_XSHAPE_FILE(p) shapexfitsname = os.path.basename(shapexfits) # Log that we are saving tilt file wmsg = 'Saving shape x information in file: {0}' WLOG(p, '', wmsg.format(shapexfitsname)) # Copy keys from fits file hdict = spirouImage.CopyOriginalKeys(fphdr) # add version number hdict = spirouImage.AddKey(p, hdict, p['KW_VERSION']) hdict = spirouImage.AddKey(p, hdict, p['KW_DRS_DATE'], value=p['DRS_DATE']) hdict = spirouImage.AddKey(p, hdict, p['KW_DATE_NOW'], value=p['DATE_NOW']) hdict = spirouImage.AddKey(p, hdict, p['KW_PID'], value=p['PID']) hdict = spirouImage.AddKey(p, hdict, p['KW_OUTPUT'], value=tag) hdict = spirouImage.AddKey(p, hdict, p['KW_CDBDARK'], value=p['DARKFILE']) hdict = spirouImage.AddKey(p, hdict, p['KW_CDBBAD'], value=p['BADPFILE']) hdict = spirouImage.AddKey(p, hdict, p['KW_CDBLOCO'], value=p['LOCOFILE']) hdict = spirouImage.AddKey1DList(p, hdict, p['KW_INFILE1'], dim1name='hcfile', values=p['HCFILE']) hdict = spirouImage.AddKey1DList(p, hdict, p['KW_INFILE2'], dim1name='fpfile', values=p['FPFILES']) # add qc parameters hdict = spirouImage.AddKey(p, hdict, p['KW_DRS_QC'], value=p['QC']) hdict = spirouImage.AddQCKeys(p, hdict, qc_params) # write tilt file to file p = spirouImage.WriteImageTable(p, shapexfits, image=loc['DXMAP'], table=fptable, hdict=hdict) # ------------------------------------------------------------------ # Writing DYMAP to file # ------------------------------------------------------------------ # get the raw tilt file name raw_shape_file = os.path.basename(p['FITSFILENAME']) # construct file name and path shapeyfits, tag = spirouConfig.Constants.SLIT_YSHAPE_FILE(p) shapeyfitsname = os.path.basename(shapeyfits) # Log that we are saving tilt file wmsg = 'Saving shape y information in file: {0}' WLOG(p, '', wmsg.format(shapeyfitsname)) # Copy keys from fits file hdict = spirouImage.CopyOriginalKeys(fphdr) # add version number hdict = spirouImage.AddKey(p, hdict, p['KW_VERSION']) hdict = spirouImage.AddKey(p, hdict, p['KW_DRS_DATE'], value=p['DRS_DATE']) hdict = spirouImage.AddKey(p, hdict, p['KW_DATE_NOW'], value=p['DATE_NOW']) hdict = spirouImage.AddKey(p, hdict, p['KW_PID'], value=p['PID']) hdict = spirouImage.AddKey(p, hdict, p['KW_OUTPUT'], value=tag) hdict = spirouImage.AddKey(p, hdict, p['KW_CDBDARK'], value=p['DARKFILE']) hdict = spirouImage.AddKey(p, hdict, p['KW_CDBBAD'], value=p['BADPFILE']) hdict = spirouImage.AddKey(p, hdict, p['KW_CDBLOCO'], value=p['LOCOFILE']) hdict = spirouImage.AddKey1DList(p, hdict, p['KW_INFILE1'], dim1name='hcfile', values=p['HCFILE']) hdict = spirouImage.AddKey1DList(p, hdict, p['KW_INFILE2'], dim1name='fpfile', values=p['FPFILES']) # add qc parameters hdict = spirouImage.AddKey(p, hdict, p['KW_DRS_QC'], value=p['QC']) hdict = spirouImage.AddQCKeys(p, hdict, qc_params) # write tilt file to file p = spirouImage.WriteImageTable(p, shapeyfits, image=loc['DYMAP'], table=fptable, hdict=hdict) # ------------------------------------------------------------------ # Writing Master FP to file # ------------------------------------------------------------------ # get the raw tilt file name raw_shape_file = os.path.basename(p['FITSFILENAME']) # construct file name and path fpmasterfits, tag = spirouConfig.Constants.SLIT_MASTER_FP_FILE(p) fpmasterfitsname = os.path.basename(fpmasterfits) # Log that we are saving tilt file wmsg = 'Saving master FP file: {0}' WLOG(p, '', wmsg.format(fpmasterfitsname)) # Copy keys from fits file hdict = spirouImage.CopyOriginalKeys(fphdr) # add version number hdict = spirouImage.AddKey(p, hdict, p['KW_VERSION']) hdict = spirouImage.AddKey(p, hdict, p['KW_DRS_DATE'], value=p['DRS_DATE']) hdict = spirouImage.AddKey(p, hdict, p['KW_DATE_NOW'], value=p['DATE_NOW']) hdict = spirouImage.AddKey(p, hdict, p['KW_PID'], value=p['PID']) hdict = spirouImage.AddKey(p, hdict, p['KW_OUTPUT'], value=tag) hdict = spirouImage.AddKey(p, hdict, p['KW_CDBDARK'], value=p['DARKFILE']) hdict = spirouImage.AddKey(p, hdict, p['KW_CDBBAD'], value=p['BADPFILE']) hdict = spirouImage.AddKey(p, hdict, p['KW_CDBLOCO'], value=p['LOCOFILE']) hdict = spirouImage.AddKey1DList(p, hdict, p['KW_INFILE1'], dim1name='hcfile', values=p['HCFILE']) hdict = spirouImage.AddKey1DList(p, hdict, p['KW_INFILE2'], dim1name='fpfile', values=p['FPFILES']) # add qc parameters hdict = spirouImage.AddKey(p, hdict, p['KW_DRS_QC'], value=p['QC']) hdict = spirouImage.AddQCKeys(p, hdict, qc_params) # write tilt file to file p = spirouImage.WriteImageTable(p, fpmasterfits, image=masterfp, table=fptable, hdict=hdict) # ------------------------------------------------------------------ # Writing sanity check files # ------------------------------------------------------------------ if p['SHAPE_DEBUG_OUTPUTS']: # log WLOG(p, '', 'Saving debug sanity check files') # construct file names input_fp_file, tag1 = spirouConfig.Constants.SLIT_SHAPE_IN_FP_FILE(p) output_fp_file, tag2 = spirouConfig.Constants.SLIT_SHAPE_OUT_FP_FILE(p) input_hc_file, tag3 = spirouConfig.Constants.SLIT_SHAPE_IN_HC_FILE(p) output_hc_file, tag4 = spirouConfig.Constants.SLIT_SHAPE_OUT_HC_FILE(p) bdxmap_file, tag5 = spirouConfig.Constants.SLIT_SHAPE_BDXMAP_FILE(p) # write input fp file hdict = spirouImage.AddKey(p, hdict, p['KW_OUTPUT'], value=tag1) p = spirouImage.WriteImage(p, input_fp_file, loc['FPDATA1'], hdict) # write output fp file hdict = spirouImage.AddKey(p, hdict, p['KW_OUTPUT'], value=tag2) p = spirouImage.WriteImage(p, output_fp_file, loc['FPDATA2'], hdict) # write input fp file hdict = spirouImage.AddKey(p, hdict, p['KW_OUTPUT'], value=tag3) p = spirouImage.WriteImage(p, input_hc_file, loc['HCDATA1'], hdict) # write output fp file hdict = spirouImage.AddKey(p, hdict, p['KW_OUTPUT'], value=tag4) p = spirouImage.WriteImage(p, output_hc_file, loc['HCDATA2'], hdict) # write overlap file hdict = spirouImage.AddKey(p, hdict, p['KW_OUTPUT'], value=tag5) p = spirouImage.WriteImage(p, bdxmap_file, loc['DXMAP0'], hdict) # ---------------------------------------------------------------------- # Move to calibDB and update calibDB # ---------------------------------------------------------------------- if p['QC']: # add shape x keydb = 'SHAPEX' # copy shape file to the calibDB folder spirouDB.PutCalibFile(p, shapexfits) # update the master calib DB file with new key spirouDB.UpdateCalibMaster(p, keydb, shapexfitsname, fphdr) # add shape y keydb = 'SHAPEY' # copy shape file to the calibDB folder spirouDB.PutCalibFile(p, shapeyfits) # update the master calib DB file with new key spirouDB.UpdateCalibMaster(p, keydb, shapeyfitsname, fphdr) # add fp master keydb = 'FPMASTER' # copy shape file to the calibDB folder spirouDB.PutCalibFile(p, fpmasterfits) # update the master calib DB file with new key spirouDB.UpdateCalibMaster(p, keydb, fpmasterfitsname, fphdr) # ---------------------------------------------------------------------- # End Message # ---------------------------------------------------------------------- p = spirouStartup.End(p) # return a copy of locally defined variables in the memory return dict(locals())
def main(night_name=None, fpfile=None, hcfiles=None): """ cal_WAVE_E2DS.py main function, if night_name and files are None uses arguments from run time i.e.: cal_DARK_spirou.py [night_directory] [fpfile] [hcfiles] :param night_name: string or None, the folder within data raw directory containing files (also reduced directory) i.e. /data/raw/20170710 would be "20170710" but /data/raw/AT5/20180409 would be "AT5/20180409" :param fpfile: string, or None, the FP file to use for arg_file_names and fitsfilename (if None assumes arg_file_names was set from run time) :param hcfiles: string, list or None, the list of HC files to use for arg_file_names and fitsfilename (if None assumes arg_file_names was set from run time) :return ll: dictionary, containing all the local variables defined in main """ # ---------------------------------------------------------------------- # Set up # ---------------------------------------------------------------------- # test files TC2 # night_name = 'AT5/AT5-12/2018-05-29_17-41-44/' # fpfile = '2279844a_fp_fp_pp_e2dsff_AB.fits' # hcfiles = ['2279845c_hc_pp_e2dsff_AB.fits'] # test files TC3 # night_name = 'TC3/AT5/AT5-12/2018-07-24_16-17-57/' # fpfile = '2294108a_pp_e2dsff_AB.fits' # hcfiles = ['2294115c_pp_e2dsff_AB.fits'] # night_name = 'TC3/AT5/AT5-12/2018-07-25_16-49-50/' # fpfile = '2294223a_pp_e2dsff_AB.fits' # hcfiles = ['2294230c_pp_e2dsff_AB.fits'] # get parameters from config files/run time args/load paths + calibdb p = spirouStartup.Begin(recipe=__NAME__) if hcfiles is None or fpfile is None: names, types = ['fpfile', 'hcfiles'], [str, str] customargs = spirouStartup.GetCustomFromRuntime(p, [0, 1], types, names, last_multi=True) else: customargs = dict(hcfiles=hcfiles, fpfile=fpfile) # get parameters from configuration files and run time arguments p = spirouStartup.LoadArguments(p, night_name, customargs=customargs, mainfitsdir='reduced', mainfitsfile='hcfiles') # ---------------------------------------------------------------------- # Construct reference filename and get fiber type # ---------------------------------------------------------------------- p, fpfitsfilename = spirouStartup.SingleFileSetup(p, filename=p['FPFILE']) fiber1 = str(p['FIBER']) p, hcfilenames = spirouStartup.MultiFileSetup(p, files=p['HCFILES']) fiber2 = str(p['FIBER']) # set the hcfilename to the first hcfilenames hcfitsfilename = hcfilenames[0] # ---------------------------------------------------------------------- # Once we have checked the e2dsfile we can load calibDB # ---------------------------------------------------------------------- # as we have custom arguments need to load the calibration database p = spirouStartup.LoadCalibDB(p) # ---------------------------------------------------------------------- # Have to check that the fibers match # ---------------------------------------------------------------------- if fiber1 == fiber2: p['FIBER'] = fiber1 fsource = __NAME__ + '/main() & spirouStartup.GetFiberType()' p.set_source('FIBER', fsource) else: emsg = 'Fiber not matching for {0} and {1}, should be the same' eargs = [hcfitsfilename, fpfitsfilename] WLOG(p, 'error', emsg.format(*eargs)) # set the fiber type p['FIB_TYP'] = [p['FIBER']] p.set_source('FIB_TYP', __NAME__ + '/main()') # ---------------------------------------------------------------------- # Read FP and HC files # ---------------------------------------------------------------------- # read and combine all HC files except the first (fpfitsfilename) rargs = [p, 'add', hcfitsfilename, hcfilenames[1:]] p, hcdata, hchdr = spirouImage.ReadImageAndCombine(*rargs) # read first file (fpfitsfilename) fpdata, fphdr, _, _ = spirouImage.ReadImage(p, fpfitsfilename) # TODO: ------------------------------------------------------------ # TODO remove to test NaNs # TODO: ------------------------------------------------------------ # hcmask = np.isfinite(hcdata) # fpmask = np.isfinite(fpdata) # hcdata[~hcmask] = 0.0 # fpdata[~fpmask] = 0.0 # TODO: ------------------------------------------------------------ # add data and hdr to loc loc = ParamDict() loc['HCDATA'], loc['HCHDR'] = hcdata, hchdr loc['FPDATA'], loc['FPHDR'] = fpdata, fphdr # set the source sources = ['HCDATA', 'HCHDR'] loc.set_sources(sources, 'spirouImage.ReadImageAndCombine()') sources = ['FPDATA', 'FPHDR'] loc.set_sources(sources, 'spirouImage.ReadImage()') # ---------------------------------------------------------------------- # Get basic image properties for reference file # ---------------------------------------------------------------------- # get sig det value p = spirouImage.GetSigdet(p, hchdr, name='sigdet') # get exposure time p = spirouImage.GetExpTime(p, hchdr, name='exptime') # get gain p = spirouImage.GetGain(p, hchdr, name='gain') # get acquisition time p = spirouImage.GetAcqTime(p, hchdr, name='acqtime', kind='julian') bjdref = p['ACQTIME'] # set sigdet and conad keywords (sigdet is changed later) p['KW_CCD_SIGDET'][1] = p['SIGDET'] p['KW_CCD_CONAD'][1] = p['GAIN'] # get lamp parameters p = spirouTHORCA.GetLampParams(p, hchdr) # get number of orders # we always get fibre A number because AB is doubled in constants file loc['NBO'] = p['QC_LOC_NBO_FPALL']['A'] loc.set_source('NBO', __NAME__ + '.main()') # get number of pixels in x from hcdata size loc['NBPIX'] = loc['HCDATA'].shape[1] loc.set_source('NBPIX', __NAME__ + '.main()') # ---------------------------------------------------------------------- # Read blaze # ---------------------------------------------------------------------- # get tilts p, loc['BLAZE'] = spirouImage.ReadBlazeFile(p, hchdr) loc.set_source('BLAZE', __NAME__ + '/main() + /spirouImage.ReadBlazeFile') # make copy of blaze (as it's overwritten later) loc['BLAZE2'] = np.copy(loc['BLAZE']) # ---------------------------------------------------------------------- # Read wave solution # ---------------------------------------------------------------------- # wavelength file; we will use the polynomial terms in its header, # NOT the pixel values that would need to be interpolated # set source of wave file wsource = __NAME__ + '/main() + /spirouImage.GetWaveSolution' # Force A and B to AB solution if p['FIBER'] in ['A', 'B']: wave_fiber = 'AB' else: wave_fiber = p['FIBER'] # get wave image wout = spirouImage.GetWaveSolution(p, hdr=hchdr, return_wavemap=True, return_filename=True, fiber=wave_fiber) loc['WAVEPARAMS'], loc['WAVE_INIT'], loc['WAVEFILE'], loc['WSOURCE'] = wout loc.set_sources(['WAVE_INIT', 'WAVEFILE', 'WAVEPARAMS', 'WSOURCE'], wsource) poly_wave_sol = loc['WAVEPARAMS'] # ---------------------------------------------------------------------- # Check that wave parameters are consistent with "ic_ll_degr_fit" # ---------------------------------------------------------------------- loc = spirouImage.CheckWaveSolConsistency(p, loc) # ---------------------------------------------------------------------- # Read UNe solution # ---------------------------------------------------------------------- wave_u_ne, amp_u_ne = spirouImage.ReadLineList(p) loc['LL_LINE'], loc['AMPL_LINE'] = wave_u_ne, amp_u_ne source = __NAME__ + '.main() + spirouImage.ReadLineList()' loc.set_sources(['ll_line', 'ampl_line'], source) # ---------------------------------------------------------------------- # Generate wave map from wave solution # ---------------------------------------------------------------------- loc = spirouWAVE.generate_wave_map(p, loc) # ---------------------------------------------------------------------- # Find Gaussian Peaks in HC spectrum # ---------------------------------------------------------------------- loc = spirouWAVE.find_hc_gauss_peaks(p, loc) # ---------------------------------------------------------------------- # Start plotting session # ---------------------------------------------------------------------- if p['DRS_PLOT'] > 0: # start interactive plot sPlt.start_interactive_session(p) # ---------------------------------------------------------------------- # Fit Gaussian peaks (in triplets) to # ---------------------------------------------------------------------- loc = spirouWAVE.fit_gaussian_triplets(p, loc) # ---------------------------------------------------------------------- # Generate Resolution map and line profiles # ---------------------------------------------------------------------- # log progress wmsg = 'Generating resolution map and ' # generate resolution map loc = spirouWAVE.generate_resolution_map(p, loc) # map line profile map if p['DRS_PLOT'] > 0: sPlt.wave_ea_plot_line_profiles(p, loc) # ---------------------------------------------------------------------- # End plotting session # ---------------------------------------------------------------------- # end interactive session if p['DRS_PLOT'] > 0: sPlt.end_interactive_session(p) # ---------------------------------------------------------------------- # Set up all_lines storage # ---------------------------------------------------------------------- # initialise up all_lines storage all_lines_1 = [] # get parameters from p n_ord_start = p['IC_HC_N_ORD_START_2'] n_ord_final = p['IC_HC_N_ORD_FINAL_2'] pixel_shift_inter = p['PIXEL_SHIFT_INTER'] pixel_shift_slope = p['PIXEL_SHIFT_SLOPE'] # get values from loc xgau = np.array(loc['XGAU_T']) dv = np.array(loc['DV_T']) fit_per_order = np.array(loc['POLY_WAVE_SOL']) ew = np.array(loc['EW_T']) peak = np.array(loc['PEAK_T']) amp_catalog = np.array(loc['AMP_CATALOG']) wave_catalog = np.array(loc['WAVE_CATALOG']) ord_t = np.array(loc['ORD_T']) # loop through orders for iord in range(n_ord_start, n_ord_final): # keep relevant lines # -> right order # -> finite dv gg = (ord_t == iord) & (np.isfinite(dv)) nlines = np.nansum(gg) # put lines into ALL_LINES structure # reminder: # gparams[0] = output wavelengths # gparams[1] = output sigma(gauss fit width) # gparams[2] = output amplitude(gauss fit) # gparams[3] = difference in input / output wavelength # gparams[4] = input amplitudes # gparams[5] = output pixel positions # gparams[6] = output pixel sigma width (gauss fit width in pixels) # gparams[7] = output weights for the pixel position chebval = np.polynomial.chebyshev.chebval # dummy array for weights test = np.ones(np.shape(xgau[gg]), 'd') * 1e4 # get the final wavelength value for each peak in order output_wave_1 = np.polyval(fit_per_order[iord][::-1], xgau[gg]) # output_wave_1 = chebval(xgau[gg], fit_per_order[iord]) # convert the pixel equivalent width to wavelength units xgau_ew_ini = xgau[gg] - ew[gg] / 2 xgau_ew_fin = xgau[gg] + ew[gg] / 2 ew_ll_ini = np.polyval(fit_per_order[iord, :], xgau_ew_ini) ew_ll_fin = np.polyval(fit_per_order[iord, :], xgau_ew_fin) # ew_ll_ini = chebval(xgau_ew_ini, fit_per_order[iord]) # ew_ll_fin = chebval(xgau_ew_fin, fit_per_order[iord]) ew_ll = ew_ll_fin - ew_ll_ini # put all lines in the order into array gau_params = np.column_stack( (output_wave_1, ew_ll, peak[gg], wave_catalog[gg] - output_wave_1, amp_catalog[gg], xgau[gg], ew[gg], test)) # append the array for the order into a list all_lines_1.append(gau_params) # save dv in km/s and auxiliary order number # res_1 = np.concatenate((res_1,2.997e5*(input_wave - output_wave_1)/ # output_wave_1)) # ord_save = np.concatenate((ord_save, test*iord)) # add to loc loc['ALL_LINES_1'] = all_lines_1 loc['LL_PARAM_1'] = np.array(fit_per_order) loc['LL_OUT_1'] = np.array(loc['WAVE_MAP2']) loc.set_sources(['ALL_LINES_1', 'LL_PARAM_1'], __NAME__ + '/main()') # For compatibility w/already defined functions, I need to save # here all_lines_2 all_lines_2 = list(all_lines_1) loc['ALL_LINES_2'] = all_lines_2 # loc['LL_PARAM_2'] = np.fliplr(fit_per_order) # loc['LL_OUT_2'] = np.array(loc['WAVE_MAP2']) # loc.set_sources(['ALL_LINES_2', 'LL_PARAM_2'], __NAME__ + '/main()') # ------------------------------------------------------------------ # Littrow test # ------------------------------------------------------------------ start = p['IC_LITTROW_ORDER_INIT_1'] end = p['IC_LITTROW_ORDER_FINAL_1'] # calculate echelle orders o_orders = np.arange(start, end) echelle_order = p['IC_HC_T_ORDER_START'] - o_orders loc['ECHELLE_ORDERS'] = echelle_order loc.set_source('ECHELLE_ORDERS', __NAME__ + '/main()') # reset Littrow fit degree p['IC_LITTROW_FIT_DEG_1'] = 7 # Do Littrow check ckwargs = dict(ll=loc['LL_OUT_1'][start:end, :], iteration=1, log=True) loc = spirouTHORCA.CalcLittrowSolution(p, loc, **ckwargs) # Plot wave solution littrow check if p['DRS_PLOT'] > 0: # plot littrow x pixels against fitted wavelength solution sPlt.wave_littrow_check_plot(p, loc, iteration=1) # ------------------------------------------------------------------ # extrapolate Littrow solution # ------------------------------------------------------------------ ekwargs = dict(ll=loc['LL_OUT_1'], iteration=1) loc = spirouTHORCA.ExtrapolateLittrowSolution(p, loc, **ekwargs) # ------------------------------------------------------------------ # Plot littrow solution # ------------------------------------------------------------------ if p['DRS_PLOT'] > 0: # plot littrow x pixels against fitted wavelength solution sPlt.wave_littrow_extrap_plot(p, loc, iteration=1) # ------------------------------------------------------------------ # Incorporate FP into solution # ------------------------------------------------------------------ # Copy LL_OUT_1 and LL_PARAM_1 into new constants (for FP integration) loc['LITTROW_EXTRAP_SOL_1'] = np.array(loc['LL_OUT_1']) loc['LITTROW_EXTRAP_PARAM_1'] = np.array(loc['LL_PARAM_1']) # only use FP if switched on in constants file if p['IC_WAVE_USE_FP']: # ------------------------------------------------------------------ # Find FP lines # ------------------------------------------------------------------ # print message to screen wmsg = 'Identification of lines in reference file: {0}' WLOG(p, '', wmsg.format(fpfile)) # ------------------------------------------------------------------ # Get the FP solution # ------------------------------------------------------------------ loc = spirouTHORCA.FPWaveSolutionNew(p, loc) # ------------------------------------------------------------------ # FP solution plots # ------------------------------------------------------------------ if p['DRS_PLOT'] > 0: # Plot the FP extracted spectrum against wavelength solution sPlt.wave_plot_final_fp_order(p, loc, iteration=1) # Plot the measured FP cavity width offset against line number sPlt.wave_local_width_offset_plot(p, loc) # Plot the FP line wavelength residuals sPlt.wave_fp_wavelength_residuals(p, loc) # ------------------------------------------------------------------ # Create new wavelength solution # ------------------------------------------------------------------ # TODO: Melissa fault - fix later p['IC_HC_N_ORD_START_2'] = min(p['IC_HC_N_ORD_START_2'], p['IC_FP_N_ORD_START']) p['IC_HC_N_ORD_FINAL_2'] = max(p['IC_HC_N_ORD_FINAL_2'], p['IC_FP_N_ORD_FINAL']) start = p['IC_HC_N_ORD_START_2'] end = p['IC_HC_N_ORD_FINAL_2'] # recalculate echelle orders for Fit1DSolution o_orders = np.arange(start, end) echelle_order = p['IC_HC_T_ORDER_START'] - o_orders loc['ECHELLE_ORDERS'] = echelle_order loc.set_source('ECHELLE_ORDERS', __NAME__ + '/main()') # select the orders to fit lls = loc['LITTROW_EXTRAP_SOL_1'][start:end] loc = spirouTHORCA.Fit1DSolution(p, loc, lls, iteration=2) # from here, LL_OUT_2 wil be 0-47 # ------------------------------------------------------------------ # Repeat Littrow test # ------------------------------------------------------------------ start = p['IC_LITTROW_ORDER_INIT_2'] end = p['IC_LITTROW_ORDER_FINAL_2'] # recalculate echelle orders for Littrow check o_orders = np.arange(start, end) echelle_order = p['IC_HC_T_ORDER_START'] - o_orders loc['ECHELLE_ORDERS'] = echelle_order loc.set_source('ECHELLE_ORDERS', __NAME__ + '/main()') # Do Littrow check ckwargs = dict(ll=loc['LL_OUT_2'][start:end, :], iteration=2, log=True) loc = spirouTHORCA.CalcLittrowSolution(p, loc, **ckwargs) # Plot wave solution littrow check if p['DRS_PLOT'] > 0: # plot littrow x pixels against fitted wavelength solution sPlt.wave_littrow_check_plot(p, loc, iteration=2) # ------------------------------------------------------------------ # extrapolate Littrow solution # ------------------------------------------------------------------ ekwargs = dict(ll=loc['LL_OUT_2'], iteration=2) loc = spirouTHORCA.ExtrapolateLittrowSolution(p, loc, **ekwargs) # ------------------------------------------------------------------ # Plot littrow solution # ------------------------------------------------------------------ if p['DRS_PLOT'] > 0: # plot littrow x pixels against fitted wavelength solution sPlt.wave_littrow_extrap_plot(p, loc, iteration=2) # ------------------------------------------------------------------ # Join 0-47 and 47-49 solutions # ------------------------------------------------------------------ loc = spirouTHORCA.JoinOrders(p, loc) # ------------------------------------------------------------------ # Plot single order, wavelength-calibrated, with found lines # ------------------------------------------------------------------ if p['DRS_PLOT'] > 0: sPlt.wave_ea_plot_single_order(p, loc) # ---------------------------------------------------------------------- # Do correlation on FP spectra # ---------------------------------------------------------------------- # ------------------------------------------------------------------ # Compute photon noise uncertainty for FP # ------------------------------------------------------------------ # set up the arguments for DeltaVrms2D dargs = [loc['FPDATA'], loc['LL_FINAL']] dkwargs = dict(sigdet=p['IC_DRIFT_NOISE'], size=p['IC_DRIFT_BOXSIZE'], threshold=p['IC_DRIFT_MAXFLUX']) # run DeltaVrms2D dvrmsref, wmeanref = spirouRV.DeltaVrms2D(*dargs, **dkwargs) # save to loc loc['DVRMSREF'], loc['WMEANREF'] = dvrmsref, wmeanref loc.set_sources(['dvrmsref', 'wmeanref'], __NAME__ + '/main()()') # log the estimated RV uncertainty wmsg = 'On fiber {0} estimated RV uncertainty on spectrum is {1:.3f} m/s' WLOG(p, 'info', wmsg.format(p['FIBER'], wmeanref)) # Use CCF Mask function with drift constants p['CCF_MASK'] = p['DRIFT_CCF_MASK'] p['TARGET_RV'] = p['DRIFT_TARGET_RV'] p['CCF_WIDTH'] = p['DRIFT_CCF_WIDTH'] p['CCF_STEP'] = p['DRIFT_CCF_STEP'] p['RVMIN'] = p['TARGET_RV'] - p['CCF_WIDTH'] p['RVMAX'] = p['TARGET_RV'] + p['CCF_WIDTH'] + p['CCF_STEP'] # get the CCF mask from file (check location of mask) loc = spirouRV.GetCCFMask(p, loc) # TODO Check why Blaze makes bugs in correlbin loc['BLAZE'] = np.ones((loc['NBO'], loc['NBPIX'])) # set sources # loc.set_sources(['flat', 'blaze'], __NAME__ + '/main()') loc.set_source('blaze', __NAME__ + '/main()') # ---------------------------------------------------------------------- # Do correlation on FP # ---------------------------------------------------------------------- # calculate and fit the CCF loc['E2DSFF'] = np.array(loc['FPDATA']) loc.set_source('E2DSFF', __NAME__ + '/main()') p['CCF_FIT_TYPE'] = 1 loc['BERV'] = 0.0 loc['BERV_MAX'] = 0.0 loc['BJD'] = 0.0 # run the RV coravelation function with these parameters loc['WAVE_LL'] = np.array(loc['LL_FINAL']) loc['PARAM_LL'] = np.array(loc['LL_PARAM_FINAL']) loc = spirouRV.Coravelation(p, loc) # ---------------------------------------------------------------------- # Update the Correlation stats with values using fiber C (FP) drift # ---------------------------------------------------------------------- # get the maximum number of orders to use nbmax = p['CCF_NUM_ORDERS_MAX'] # get the average ccf loc['AVERAGE_CCF'] = np.nansum(loc['CCF'][:nbmax], axis=0) # normalize the average ccf normalized_ccf = loc['AVERAGE_CCF'] / np.nanmax(loc['AVERAGE_CCF']) # get the fit for the normalized average ccf ccf_res, ccf_fit = spirouRV.FitCCF(p, loc['RV_CCF'], normalized_ccf, fit_type=1) loc['CCF_RES'] = ccf_res loc['CCF_FIT'] = ccf_fit # get the max cpp loc['MAXCPP'] = np.nansum(loc['CCF_MAX']) / np.nansum( loc['PIX_PASSED_ALL']) # get the RV value from the normalised average ccf fit center location loc['RV'] = float(ccf_res[1]) # get the contrast (ccf fit amplitude) loc['CONTRAST'] = np.abs(100 * ccf_res[0]) # get the FWHM value loc['FWHM'] = ccf_res[2] * spirouCore.spirouMath.fwhm() # set the source keys = [ 'AVERAGE_CCF', 'MAXCPP', 'RV', 'CONTRAST', 'FWHM', 'CCF_RES', 'CCF_FIT' ] loc.set_sources(keys, __NAME__ + '/main()') # ---------------------------------------------------------------------- # log the stats wmsg = ('FP Correlation: C={0:.1f}[%] DRIFT={1:.5f}[km/s] ' 'FWHM={2:.4f}[km/s] maxcpp={3:.1f}') wargs = [loc['CONTRAST'], float(ccf_res[1]), loc['FWHM'], loc['MAXCPP']] WLOG(p, 'info', wmsg.format(*wargs)) # ---------------------------------------------------------------------- # rv ccf plot # ---------------------------------------------------------------------- if p['DRS_PLOT'] > 0: # Plot rv vs ccf (and rv vs ccf_fit) p['OBJNAME'] = 'FP' sPlt.ccf_rv_ccf_plot(p, loc['RV_CCF'], normalized_ccf, ccf_fit) # TODO : Add QC of the FP CCF # ---------------------------------------------------------------------- # Quality control # ---------------------------------------------------------------------- # get parameters ffrom p p['QC_RMS_LITTROW_MAX'] = p['QC_HC_RMS_LITTROW_MAX'] p['QC_DEV_LITTROW_MAX'] = p['QC_HC_DEV_LITTROW_MAX'] # set passed variable and fail message list passed, fail_msg = True, [] qc_values, qc_names, qc_logic, qc_pass = [], [], [], [] # ---------------------------------------------------------------------- # quality control on sigma clip (sig1 > qc_hc_wave_sigma_max if loc['SIG1'] > p['QC_HC_WAVE_SIGMA_MAX']: fmsg = 'Sigma too high ({0:.5f} > {1:.5f})' fail_msg.append(fmsg.format(loc['SIG1'], p['QC_HC_WAVE_SIGMA_MAX'])) passed = False qc_pass.append(0) else: qc_pass.append(1) # add to qc header lists qc_values.append(loc['SIG1']) qc_names.append('SIG1') qc_logic.append('SIG1 > {0:.2f}'.format(p['QC_HC_WAVE_SIGMA_MAX'])) # ---------------------------------------------------------------------- # check the difference between consecutive orders is always positive # get the differences wave_diff = loc['LL_FINAL'][1:] - loc['LL_FINAL'][:-1] if np.min(wave_diff) < 0: fmsg = 'Negative wavelength difference between orders' fail_msg.append(fmsg) passed = False qc_pass.append(0) else: qc_pass.append(1) # add to qc header lists qc_values.append(np.min(wave_diff)) qc_names.append('MIN WAVE DIFF') qc_logic.append('MIN WAVE DIFF < 0') # ---------------------------------------------------------------------- # check for infinites and NaNs in mean residuals from fit if ~np.isfinite(loc['X_MEAN_2']): # add failed message to the fail message list fmsg = 'NaN or Inf in X_MEAN_2' fail_msg.append(fmsg) passed = False qc_pass.append(0) else: qc_pass.append(1) # add to qc header lists qc_values.append(loc['X_MEAN_2']) qc_names.append('X_MEAN_2') qc_logic.append('X_MEAN_2 not finite') # ---------------------------------------------------------------------- # iterate through Littrow test cut values lit_it = 2 # checks every other value # TODO: This QC check (or set of QC checks needs re-writing it is # TODO: nearly impossible to understand for x_it in range(1, len(loc['X_CUT_POINTS_' + str(lit_it)]), 2): # get x cut point x_cut_point = loc['X_CUT_POINTS_' + str(lit_it)][x_it] # get the sigma for this cut point sig_littrow = loc['LITTROW_SIG_' + str(lit_it)][x_it] # get the abs min and max dev littrow values min_littrow = abs(loc['LITTROW_MINDEV_' + str(lit_it)][x_it]) max_littrow = abs(loc['LITTROW_MAXDEV_' + str(lit_it)][x_it]) # get the corresponding order min_littrow_ord = loc['LITTROW_MINDEVORD_' + str(lit_it)][x_it] max_littrow_ord = loc['LITTROW_MAXDEVORD_' + str(lit_it)][x_it] # check if sig littrow is above maximum rms_littrow_max = p['QC_RMS_LITTROW_MAX'] dev_littrow_max = p['QC_DEV_LITTROW_MAX'] if sig_littrow > rms_littrow_max: fmsg = ('Littrow test (x={0}) failed (sig littrow = ' '{1:.2f} > {2:.2f})') fargs = [x_cut_point, sig_littrow, rms_littrow_max] fail_msg.append(fmsg.format(*fargs)) passed = False qc_pass.append(0) else: qc_pass.append(1) # add to qc header lists qc_values.append(sig_littrow) qc_names.append('sig_littrow') qc_logic.append('sig_littrow > {0:.2f}'.format(rms_littrow_max)) # ---------------------------------------------------------------------- # check if min/max littrow is out of bounds if np.max([max_littrow, min_littrow]) > dev_littrow_max: fmsg = ('Littrow test (x={0}) failed (min|max dev = ' '{1:.2f}|{2:.2f} > {3:.2f} for order {4}|{5})') fargs = [ x_cut_point, min_littrow, max_littrow, dev_littrow_max, min_littrow_ord, max_littrow_ord ] fail_msg.append(fmsg.format(*fargs)) passed = False qc_pass.append(0) # TODO: Should this be the QC header values? # TODO: it does not change the outcome of QC (i.e. passed=False) # TODO: So what is the point? # if sig was out of bounds, recalculate if sig_littrow > rms_littrow_max: # conditions check1 = min_littrow > dev_littrow_max check2 = max_littrow > dev_littrow_max # get the residuals respix = loc['LITTROW_YY_' + str(lit_it)][x_it] # check if both are out of bounds if check1 and check2: # remove respective orders worst_order = (min_littrow_ord, max_littrow_ord) respix_2 = np.delete(respix, worst_order) redo_sigma = True # check if min is out of bounds elif check1: # remove respective order worst_order = min_littrow_ord respix_2 = np.delete(respix, worst_order) redo_sigma = True # check if max is out of bounds elif check2: # remove respective order worst_order = max_littrow_ord respix_2 = np.delete(respix, max_littrow_ord) redo_sigma = True # else do not recalculate sigma else: redo_sigma, respix_2, worst_order = False, None, None wmsg = 'No outlying orders, sig littrow not recalculated' fail_msg.append(wmsg.format()) # if outlying order, recalculate stats if redo_sigma: mean = np.nansum(respix_2) / len(respix_2) mean2 = np.nansum(respix_2**2) / len(respix_2) rms = np.sqrt(mean2 - mean**2) if rms > rms_littrow_max: fmsg = ('Littrow test (x={0}) failed (sig littrow = ' '{1:.2f} > {2:.2f} removing order {3})') fargs = [ x_cut_point, rms, rms_littrow_max, worst_order ] fail_msg.append(fmsg.format(*fargs)) else: wargs = [ x_cut_point, rms, rms_littrow_max, worst_order ] wmsg = ('Littrow test (x={0}) passed (sig littrow = ' '{1:.2f} > {2:.2f} removing order {3})') fail_msg.append(wmsg.format(*wargs)) else: qc_pass.append(1) # add to qc header lists qc_values.append(np.max([max_littrow, min_littrow])) qc_names.append('max or min littrow') qc_logic.append('max or min littrow > {0:.2f}' ''.format(dev_littrow_max)) # finally log the failed messages and set QC = 1 if we pass the # quality control QC = 0 if we fail quality control if passed: WLOG(p, 'info', 'QUALITY CONTROL SUCCESSFUL - Well Done -') p['QC'] = 1 p.set_source('QC', __NAME__ + '/main()') else: for farg in fail_msg: wmsg = 'QUALITY CONTROL FAILED: {0}' WLOG(p, 'warning', wmsg.format(farg)) p['QC'] = 0 p.set_source('QC', __NAME__ + '/main()') # store in qc_params qc_params = [qc_names, qc_values, qc_logic, qc_pass] # ------------------------------------------------------------------ # archive result in e2ds spectra # ------------------------------------------------------------------ # get raw input file name(s) raw_infiles1 = [] for hcfile in p['HCFILES']: raw_infiles1.append(os.path.basename(hcfile)) raw_infile2 = os.path.basename(p['FPFILE']) # get wave filename wavefits, tag1 = spirouConfig.Constants.WAVE_FILE_EA_2(p) wavefitsname = os.path.split(wavefits)[-1] # log progress wargs = [p['FIBER'], wavefits] wmsg = 'Write wavelength solution for Fiber {0} in {1}' WLOG(p, '', wmsg.format(*wargs)) # write solution to fitsfilename header # copy original keys hdict = spirouImage.CopyOriginalKeys(loc['HCHDR']) # add version number hdict = spirouImage.AddKey(p, hdict, p['KW_VERSION']) hdict = spirouImage.AddKey(p, hdict, p['KW_DRS_DATE'], value=p['DRS_DATE']) hdict = spirouImage.AddKey(p, hdict, p['KW_DATE_NOW'], value=p['DATE_NOW']) hdict = spirouImage.AddKey(p, hdict, p['KW_PID'], value=p['PID']) # set the input files hdict = spirouImage.AddKey(p, hdict, p['KW_CDBBLAZE'], value=p['BLAZFILE']) hdict = spirouImage.AddKey(p, hdict, p['KW_CDBWAVE'], value=loc['WAVEFILE']) hdict = spirouImage.AddKey(p, hdict, p['KW_WAVESOURCE'], value=loc['WSOURCE']) hdict = spirouImage.AddKey1DList(p, hdict, p['KW_INFILE1'], dim1name='fpfile', values=p['FPFILE']) hdict = spirouImage.AddKey1DList(p, hdict, p['KW_INFILE2'], dim1name='hcfile', values=p['HCFILES']) # add qc parameters hdict = spirouImage.AddKey(p, hdict, p['KW_DRS_QC'], value=p['QC']) hdict = spirouImage.AddQCKeys(p, hdict, qc_params) # add wave solution date hdict = spirouImage.AddKey(p, hdict, p['KW_WAVE_TIME1'], value=p['MAX_TIME_HUMAN']) hdict = spirouImage.AddKey(p, hdict, p['KW_WAVE_TIME2'], value=p['MAX_TIME_UNIX']) hdict = spirouImage.AddKey(p, hdict, p['KW_WAVE_CODE'], value=__NAME__) # add number of orders hdict = spirouImage.AddKey(p, hdict, p['KW_WAVE_ORD_N'], value=loc['LL_PARAM_FINAL'].shape[0]) # add degree of fit hdict = spirouImage.AddKey(p, hdict, p['KW_WAVE_LL_DEG'], value=loc['LL_PARAM_FINAL'].shape[1] - 1) # add wave solution hdict = spirouImage.AddKey2DList(p, hdict, p['KW_WAVE_PARAM'], values=loc['LL_PARAM_FINAL']) # add FP CCF drift # target RV and width hdict = spirouImage.AddKey(p, hdict, p['KW_WFP_TARG_RV'], value=p['TARGET_RV']) hdict = spirouImage.AddKey(p, hdict, p['KW_WFP_WIDTH'], value=p['CCF_WIDTH']) # the rv step # rvstep = np.abs(loc['RV_CCF'][0] - loc['RV_CCF'][1]) # hdict = spirouImage.AddKey(p, hdict, p['KW_CCF_CDELT'], value=rvstep) hdict = spirouImage.AddKey(p, hdict, p['KW_WFP_STEP'], value=p['CCF_STEP']) # add ccf stats hdict = spirouImage.AddKey(p, hdict, p['KW_WFP_DRIFT'], value=loc['CCF_RES'][1]) hdict = spirouImage.AddKey(p, hdict, p['KW_WFP_FWHM'], value=loc['FWHM']) hdict = spirouImage.AddKey(p, hdict, p['KW_WFP_CONTRAST'], value=loc['CONTRAST']) hdict = spirouImage.AddKey(p, hdict, p['KW_WFP_MAXCPP'], value=loc['MAXCPP']) hdict = spirouImage.AddKey(p, hdict, p['KW_WFP_MASK'], value=p['CCF_MASK']) hdict = spirouImage.AddKey(p, hdict, p['KW_WFP_LINES'], value=np.nansum(loc['TOT_LINE'])) # write the wave "spectrum" hdict = spirouImage.AddKey(p, hdict, p['KW_OUTPUT'], value=tag1) p = spirouImage.WriteImage(p, wavefits, loc['LL_FINAL'], hdict) # get filename for E2DS calibDB copy of FITSFILENAME e2dscopy_filename = spirouConfig.Constants.WAVE_E2DS_COPY(p)[0] wargs = [p['FIBER'], os.path.split(e2dscopy_filename)[-1]] wmsg = 'Write reference E2DS spectra for Fiber {0} in {1}' WLOG(p, '', wmsg.format(*wargs)) # make a copy of the E2DS file for the calibBD p = spirouImage.WriteImage(p, e2dscopy_filename, loc['HCDATA'], hdict) # only copy over if QC passed if p['QC']: # loop around hc files and update header with for hcfile in p['HCFILES']: raw_infilepath1 = os.path.join(p['ARG_FILE_DIR'], hcfile) p = spirouImage.UpdateWaveSolution(p, loc, raw_infilepath1) # update fp file raw_infilepath2 = os.path.join(p['ARG_FILE_DIR'], raw_infile2) p = spirouImage.UpdateWaveSolution(p, loc, raw_infilepath2) # ------------------------------------------------------------------ # Save to result table # ------------------------------------------------------------------ # calculate stats for table final_mean = 1000 * loc['X_MEAN_2'] final_var = 1000 * loc['X_VAR_2'] num_lines = int(np.nansum(loc['X_ITER_2'][:, 2])) # loc['X_ITER_2'] err = 1000 * np.sqrt(loc['X_VAR_2'] / num_lines) sig_littrow = 1000 * np.array(loc['LITTROW_SIG_' + str(lit_it)]) # construct filename wavetbl = spirouConfig.Constants.WAVE_TBL_FILE_EA(p) wavetblname = os.path.basename(wavetbl) # construct and write table columnnames = [ 'night_name', 'file_name', 'fiber', 'mean', 'rms', 'N_lines', 'err', 'rms_L500', 'rms_L1000', 'rms_L1500', 'rms_L2000', 'rms_L2500', 'rms_L3000', 'rms_L3500' ] columnformats = [ '{:20s}', '{:30s}', '{:3s}', '{:7.4f}', '{:6.2f}', '{:3d}', '{:6.3f}', '{:6.2f}', '{:6.2f}', '{:6.2f}', '{:6.2f}', '{:6.2f}', '{:6.2f}', '{:6.2f}' ] columnvalues = [[p['ARG_NIGHT_NAME']], [p['ARG_FILE_NAMES'][0]], [p['FIBER']], [final_mean], [final_var], [num_lines], [err], [sig_littrow[0]], [sig_littrow[1]], [sig_littrow[2]], [sig_littrow[3]], [sig_littrow[4]], [sig_littrow[5]], [sig_littrow[6]]] # make table table = spirouImage.MakeTable(p, columns=columnnames, values=columnvalues, formats=columnformats) # merge table wmsg = 'Global result summary saved in {0}' WLOG(p, '', wmsg.format(wavetblname)) spirouImage.MergeTable(p, table, wavetbl, fmt='ascii.rst') # ---------------------------------------------------------------------- # Save resolution and line profiles to file # ---------------------------------------------------------------------- raw_infile = os.path.basename(p['FITSFILENAME']) # get wave filename resfits, tag3 = spirouConfig.Constants.WAVE_RES_FILE_EA(p) resfitsname = os.path.basename(resfits) WLOG(p, '', 'Saving wave resmap to {0}'.format(resfitsname)) # make a copy of the E2DS file for the calibBD # set the version hdict = spirouImage.AddKey(p, hdict, p['KW_VERSION']) hdict = spirouImage.AddKey(p, hdict, p['KW_DRS_DATE'], value=p['DRS_DATE']) hdict = spirouImage.AddKey(p, hdict, p['KW_DATE_NOW'], value=p['DATE_NOW']) hdict = spirouImage.AddKey(p, hdict, p['KW_OUTPUT'], value=tag3) # get res data in correct format resdata, hdicts = spirouTHORCA.GenerateResFiles(p, loc, hdict) # save to file p = spirouImage.WriteImageMulti(p, resfits, resdata, hdicts=hdicts) # ------------------------------------------------------------------ # Save line list table file # ------------------------------------------------------------------ # construct filename # TODO proper column values wavelltbl = spirouConfig.Constants.WAVE_LINE_FILE_EA(p) wavelltblname = os.path.split(wavelltbl)[-1] # construct and write table columnnames = ['order', 'll', 'dv', 'w', 'xi', 'xo', 'dvdx'] columnformats = [ '{:.0f}', '{:12.4f}', '{:13.5f}', '{:12.4f}', '{:12.4f}', '{:12.4f}', '{:8.4f}' ] columnvalues = [] # construct column values (flatten over orders) for it in range(len(loc['X_DETAILS_2'])): for jt in range(len(loc['X_DETAILS_2'][it][0])): row = [ float(it), loc['X_DETAILS_2'][it][0][jt], loc['LL_DETAILS_2'][it][0][jt], loc['X_DETAILS_2'][it][3][jt], loc['X_DETAILS_2'][it][1][jt], loc['X_DETAILS_2'][it][2][jt], loc['SCALE_2'][it][jt] ] columnvalues.append(row) # log saving wmsg = 'List of lines used saved in {0}' WLOG(p, '', wmsg.format(wavelltblname)) # make table columnvalues = np.array(columnvalues).T table = spirouImage.MakeTable(p, columns=columnnames, values=columnvalues, formats=columnformats) # write table spirouImage.WriteTable(p, table, wavelltbl, fmt='ascii.rst') # ------------------------------------------------------------------ # Move to calibDB and update calibDB # ------------------------------------------------------------------ if p['QC']: # set the wave key keydb = 'WAVE_{0}'.format(p['FIBER']) # copy wave file to calibDB folder spirouDB.PutCalibFile(p, wavefits) # update the master calib DB file with new key spirouDB.UpdateCalibMaster(p, keydb, wavefitsname, loc['HCHDR']) # set the hcref key keydb = 'HCREF_{0}'.format(p['FIBER']) # copy wave file to calibDB folder spirouDB.PutCalibFile(p, e2dscopy_filename) # update the master calib DB file with new key e2dscopyfits = os.path.split(e2dscopy_filename)[-1] spirouDB.UpdateCalibMaster(p, keydb, e2dscopyfits, loc['HCHDR']) # ---------------------------------------------------------------------- # End Message # ---------------------------------------------------------------------- p = spirouStartup.End(p) # return p and loc return dict(locals())
def main(night_name=None, files=None): """ cal_HC_E2DS.py main function, if night_name and files are None uses arguments from run time i.e.: cal_DARK_spirou.py [night_directory] [fitsfilename] :param night_name: string or None, the folder within data raw directory containing files (also reduced directory) i.e. /data/raw/20170710 would be "20170710" but /data/raw/AT5/20180409 would be "AT5/20180409" :param files: string, list or None, the list of files to use for arg_file_names and fitsfilename (if None assumes arg_file_names was set from run time) :return ll: dictionary, containing all the local variables defined in main """ # ---------------------------------------------------------------------- # Set up # ---------------------------------------------------------------------- # get parameters from config files/run time args/load paths + calibdb p = spirouStartup.Begin(recipe=__NAME__) # get parameters from configuration files and run time arguments p = spirouStartup.LoadArguments(p, night_name, files, mainfitsdir='reduced') # setup files and get fiber p = spirouStartup.InitialFileSetup(p, calibdb=True) # set the fiber type p['FIB_TYP'] = [p['FIBER']] p.set_source('FIB_TYP', __NAME__ + '/main()') # set find line mode find_lines_mode = p['HC_FIND_LINES_MODE'] # ---------------------------------------------------------------------- # Read image file # ---------------------------------------------------------------------- # read and combine all files p, hcdata, hchdr = spirouImage.ReadImageAndCombine(p, 'add') # add data and hdr to loc loc = ParamDict() loc['HCDATA'], loc['HCHDR'] = hcdata, hchdr # set the source sources = ['HCDATA', 'HCHDR'] loc.set_sources(sources, 'spirouImage.ReadImageAndCombine()') # ---------------------------------------------------------------------- # Get basic parameters # ---------------------------------------------------------------------- # get sig det value p = spirouImage.GetSigdet(p, loc['HCHDR'], name='sigdet') # get exposure time p = spirouImage.GetExpTime(p, loc['HCHDR'], name='exptime') # get gain p = spirouImage.GetGain(p, loc['HCHDR'], name='gain') # get acquisition time p = spirouImage.GetAcqTime(p, loc['HCHDR'], name='acqtime', kind='julian') bjdref = p['ACQTIME'] # set sigdet and conad keywords (sigdet is changed later) p['KW_CCD_SIGDET'][1] = p['SIGDET'] p['KW_CCD_CONAD'][1] = p['GAIN'] # get lamp parameters p = spirouTHORCA.GetLampParams(p, loc['HCHDR']) # ---------------------------------------------------------------------- # Obtain the flat # ---------------------------------------------------------------------- # get the flat # p, loc = spirouFLAT.GetFlat(p, loc, hchdr) # correct the data with the flat # TODO: Should this be used? # log # WLOG(p, '', 'Applying flat correction') # loc['HCDATA'] = loc['HCDATA']/loc['FLAT'] # ---------------------------------------------------------------------- # Read blaze # ---------------------------------------------------------------------- # get tilts loc['BLAZE'] = spirouImage.ReadBlazeFile(p, hchdr) loc.set_source('BLAZE', __NAME__ + '/main() + /spirouImage.ReadBlazeFile') # ---------------------------------------------------------------------- # Start plotting session # ---------------------------------------------------------------------- if p['DRS_PLOT'] > 0: # start interactive plot sPlt.start_interactive_session(p) # ---------------------------------------------------------------------- # loop around fiber type # ---------------------------------------------------------------------- for fiber in p['FIB_TYP']: # set fiber type for inside loop p['FIBER'] = fiber # ------------------------------------------------------------------ # Wave solution # ------------------------------------------------------------------ # log message for loop wmsg = 'Processing Wavelength Calibration for Fiber {0}' WLOG(p, 'info', wmsg.format(p['FIBER'])) # ------------------------------------------------------------------ # Part 1 # ------------------------------------------------------------------ p, loc = part1(p, loc, mode=find_lines_mode) # ------------------------------------------------------------------ # Part 2 # ------------------------------------------------------------------ # set params for part2 p['QC_RMS_LITTROW_MAX'] = p['QC_HC_RMS_LITTROW_MAX'] p['QC_DEV_LITTROW_MAX'] = p['QC_HC_DEV_LITTROW_MAX'] # ------------------------------------------------------------------ # run part 2 p, loc = part2(p, loc) # ---------------------------------------------------------------------- # End plotting session # ---------------------------------------------------------------------- # end interactive session if p['DRS_PLOT'] > 0: sPlt.end_interactive_session(p) # ---------------------------------------------------------------------- # End Message # ---------------------------------------------------------------------- p = spirouStartup.End(p) # return a copy of locally defined variables in the memory return dict(locals())