ObservationSim.py

import os
import numpy as np
import mpi4py.MPI as MPI
import galsim
import logging
import psutil
import gc
from astropy.io import fits
from datetime import datetime

import traceback

from ObservationSim.Config import config_dir, ChipOutput
from ObservationSim.Config.Header import generatePrimaryHeader, generateExtensionHeader
from ObservationSim.Instrument import Telescope, Filter, FilterParam, FocalPlane, Chip
from ObservationSim.Instrument.Chip import Effects
from ObservationSim.MockObject import calculateSkyMap_split_g
from ObservationSim.PSF import PSFGauss, FieldDistortion, PSFInterp
from ObservationSim._util import get_shear_field, makeSubDir_PointingList
from ObservationSim.Astrometry.Astrometry_util import on_orbit_obs_position

# import tracemalloc

class Observation(object):
    def __init__(self, config, Catalog, work_dir=None, data_dir=None):
        self.path_dict = config_dir(config=config, work_dir=work_dir, data_dir=data_dir)
        self.config = config
        self.tel = Telescope()
        self.focal_plane = FocalPlane(survey_type=self.config["obs_setting"]["survey_type"]) 
        self.filter_param = FilterParam() 
        self.chip_list = []
        self.filter_list = []
        self.all_filter = []
        self.Catalog = Catalog

        # if we want to apply field distortion?
        if self.config["ins_effects"]["field_dist"] == True:
            self.fd_model = FieldDistortion(fdModel_path=self.path_dict["fd_path"])
        else:
            self.fd_model = None

        # Construct chips & filters:
        nchips = self.focal_plane.nchip_x*self.focal_plane.nchip_y
        for i in range(nchips):
            chipID = i + 1

            # Make Chip & Filter lists
            chip = Chip(
                chipID=chipID, 
                config=self.config)
            filter_id, filter_type = chip.getChipFilter()
            filt = Filter(filter_id=filter_id, 
                filter_type=filter_type, 
                filter_param=self.filter_param)
            if not self.focal_plane.isIgnored(chipID=chipID):
                self.chip_list.append(chip)
                self.filter_list.append(filt)
            self.all_filter.append(filt)

        # Read catalog and shear(s)
        self.g1_field, self.g2_field, self.nshear = get_shear_field(config=self.config)

    def run_one_chip(self, chip, filt, pointing, chip_output, wcs_fp=None, psf_model=None, shear_cat_file=None, cat_dir=None, sed_dir=None):

        # print(':::::::::::::::::::Current Pointing Information::::::::::::::::::')
        # print("RA: %f, DEC; %f" % (pointing.ra, pointing.dec))
        # print("Time: %s" % datetime.utcfromtimestamp(pointing.timestamp).isoformat())
        # print("Exposure time: %f" % pointing.exp_time)
        # print("Satellite Position (x, y, z): (%f, %f, %f)" % (pointing.sat_x, pointing.sat_y, pointing.sat_z))
        # print("Satellite Velocity (x, y, z): (%f, %f, %f)" % (pointing.sat_vx, pointing.sat_vy, pointing.sat_vz))
        # print("Position Angle: %f" % pointing.img_pa.deg)
        # print('Chip : %d' % chip.chipID)
        # print(':::::::::::::::::::::::::::END:::::::::::::::::::::::::::::::::::')
        chip_output.logger.info(':::::::::::::::::::Current Pointing Information::::::::::::::::::')
        chip_output.logger.info("RA: %f, DEC; %f" % (pointing.ra, pointing.dec))
        chip_output.logger.info("Time: %s" % datetime.utcfromtimestamp(pointing.timestamp).isoformat())
        chip_output.logger.info("Exposure time: %f" % pointing.exp_time)
        chip_output.logger.info("Satellite Position (x, y, z): (%f, %f, %f)" % (pointing.sat_x, pointing.sat_y, pointing.sat_z))
        chip_output.logger.info("Satellite Velocity (x, y, z): (%f, %f, %f)" % (pointing.sat_vx, pointing.sat_vy, pointing.sat_vz))
        chip_output.logger.info("Position Angle: %f" % pointing.img_pa.deg)
        chip_output.logger.info('Chip : %d' % chip.chipID)
        chip_output.logger.info(':::::::::::::::::::::::::::END:::::::::::::::::::::::::::::::::::')

        if self.config["psf_setting"]["psf_model"] == "Gauss":
            psf_model = PSFGauss(chip=chip, psfRa=self.config["psf_setting"]["psf_rcont"])
        elif self.config["psf_setting"]["psf_model"] == "Interp":
            psf_model = PSFInterp(chip=chip, PSF_data_file=self.path_dict["psf_dir"])
        else:
            # print("unrecognized PSF model type!!", flush=True)
            chip_output.logger.error("unrecognized PSF model type!!", flush=True)

        # Get (extra) shear fields
        if shear_cat_file is not None:
            self.g1_field, self.g2_field, self.nshear = get_shear_field(config=self.config, shear_cat_file=shear_cat_file)

        # Apply astrometric simulation for pointing
        if self.config["obs_setting"]["enable_astrometric_model"]:
            dt = datetime.utcfromtimestamp(pointing.timestamp)
            date_str = dt.date().isoformat()
            time_str = dt.time().isoformat()
            ra_cen, dec_cen = on_orbit_obs_position(
                input_ra_list=[pointing.ra],
                input_dec_list=[pointing.dec],
                input_pmra_list=[0.],
                input_pmdec_list=[0.],
                input_rv_list=[0.],
                input_parallax_list=[1e-9],
                input_nstars=1,
                input_x=pointing.sat_x,
                input_y=pointing.sat_y,
                input_z=pointing.sat_z,
                input_vx=pointing.sat_vx,
                input_vy=pointing.sat_vy,
                input_vz=pointing.sat_vz,
                input_epoch="J2000",
                input_date_str=date_str,
                input_time_str=time_str
            )
            ra_cen, dec_cen = ra_cen[0], dec_cen[0]
        else:
            ra_cen = pointing.ra
            dec_cen = pointing.dec

        # Get WCS for the focal plane
        if wcs_fp == None:
            wcs_fp = self.focal_plane.getTanWCS(ra_cen, dec_cen, pointing.img_pa, chip.pix_scale)

        # Create chip Image
        chip.img = galsim.ImageF(chip.npix_x, chip.npix_y)
        chip.img.setOrigin(chip.bound.xmin, chip.bound.ymin)
        chip.img.wcs = wcs_fp
        if chip.survey_type == "photometric":
            sky_map = None
        # elif chip.survey_type == "spectroscopic":
        #     sky_map = calculateSkyMap_split_g(xLen=chip.npix_x, yLen=chip.npix_y, blueLimit=filt.blue_limit, redLimit=filt.red_limit, skyfn=self.path_dict["sky_file"], conf=chip.sls_conf, pixelSize=chip.pix_scale, isAlongY=0)
        elif chip.survey_type == "spectroscopic":
            # chip.loadSLSFLATCUBE(flat_fn='flat_cube.fits')
            flat_normal = np.ones_like(chip.img.array)
            if self.config["ins_effects"]["flat_fielding"] == True:
                # print("SLS flat preprocess,CHIP %d : Creating and applying Flat-Fielding"%chip.chipID, flush=True)
                # print(chip.img.bounds, flush=True)
                chip_output.logger.info("SLS flat preprocess,CHIP %d : Creating and applying Flat-Fielding"%chip.chipID)
                msg = str(chip.img.bounds)
                chip_output.logger.info(msg)
                flat_img = Effects.MakeFlatSmooth(
                    chip.img.bounds,
                    int(self.config["random_seeds"]["seed_flat"]))
                flat_normal = flat_normal * flat_img.array / np.mean(flat_img.array)
            if self.config["ins_effects"]["shutter_effect"] == True:
                # print("SLS flat preprocess,CHIP %d : Apply shutter effect"%chip.chipID, flush=True)
                chip_output.logger.info("SLS flat preprocess,CHIP %d : Apply shutter effect"%chip.chipID)
                shuttimg = Effects.ShutterEffectArr(chip.img, t_shutter=1.3, dist_bearing=735,
                                                    dt=1E-3)  # shutter effect normalized image for this chip
                flat_normal = flat_normal*shuttimg
                flat_normal = np.array(flat_normal,dtype='float32')
            sky_map = calculateSkyMap_split_g(
                skyMap=flat_normal,
                blueLimit=filt.blue_limit,
                redLimit=filt.red_limit,
                conf=chip.sls_conf,
                pixelSize=chip.pix_scale,
                isAlongY=0,
                flat_cube=chip.flat_cube)
            # sky_map = np.ones([9216, 9232])
            del flat_normal

        if pointing.pointing_type == 'MS':
            # Load catalogues and templates
            self.cat = self.Catalog(config=self.config, chip=chip, pointing=pointing, cat_dir=cat_dir, sed_dir=sed_dir, chip_output=chip_output, filt=filt)
            chip_output.create_output_file()
            self.nobj = len(self.cat.objs)

            for ifilt in range(len(self.all_filter)):
                temp_filter = self.all_filter[ifilt]
                # Update the limiting magnitude using exposure time in pointing
                temp_filter.update_limit_saturation_mags(exptime=pointing.exp_time, chip=chip)

                # Select cutting band filter for saturation/limiting magnitude
                if temp_filter.filter_type.lower() == self.config["obs_setting"]["cut_in_band"].lower():
                    cut_filter = temp_filter

            # Loop over objects
            missed_obj = 0
            bright_obj = 0
            dim_obj = 0
            for j in range(self.nobj):
                
                # (DEBUG)
                # if j >= 10:
                #     break
                # tracemalloc.start()

                obj = self.cat.objs[j]

                #if obj.type == 'star' and self.config["run_option"]["galaxy_only"]:
                #    continue
                #elif obj.type == 'galaxy' and self.config["run_option"]["star_only"]:
                #    continue
                #elif obj.type == 'quasar' and self.config["run_option"]["star_only"]:
                #    continue

                ###mock_stamp_START
                if obj.type == 'star' and not self.config["run_option"]["star_yes"]:
                    continue
                elif obj.type == 'galaxy' and not self.config["run_option"]["galaxy_yes"]:
                    continue
                elif obj.type == 'quasar' and not self.config["run_option"]["galaxy_yes"]:
                    continue
                elif obj.type == 'stamp' and not self.config["run_option"]["stamp_yes"]:
                    continue
                #if obj.type == 'star' or obj.type == 'galaxy':
                #    continue #for test

                #mu_temp = 1.
                #if obj.type == 'stamp':
                #    mu_temp = obj.param["mu"]
                ###mock_stamp_END

                # load and convert SED; also caculate object's magnitude in all CSST bands
                try:
                    sed_data = self.cat.load_sed(obj)
                    norm_filt = self.cat.load_norm_filt(obj)
                    obj.sed, obj.param["mag_%s"%filt.filter_type], obj.param["flux_%s"%filt.filter_type] = self.cat.convert_sed(
                        mag=obj.param["mag_use_normal"],
                        sed=sed_data,
                        target_filt=filt, 
                        norm_filt=norm_filt,
                    )
                    _, obj.param["mag_%s"%cut_filter.filter_type], obj.param["flux_%s"%cut_filter.filter_type] = self.cat.convert_sed(
                        mag=obj.param["mag_use_normal"],
                        sed=sed_data,
                        target_filt=cut_filter, 
                        norm_filt=norm_filt,
                    )
                except Exception as e:
                    # print(e)
                    traceback.print_exc()
                    chip_output.logger.error(e)
                    continue
                # print("mag_%s = %.3f"%(filt.filter_type, obj.param["mag_%s"%filt.filter_type]))
                # chip_output.logger.info("mag_%s = %.3f"%(filt.filter_type, obj.param["mag_%s"%filt.filter_type]))

                # Exclude very bright/dim objects (for now)
                # if filt.is_too_bright(mag=obj.getMagFilter(filt)):
                # if filt.is_too_bright(mag=obj.mag_use_normal):
                if cut_filter.is_too_bright(
                    mag=obj.param["mag_%s"%self.config["obs_setting"]["cut_in_band"].lower()],
                    margin=self.config["obs_setting"]["mag_sat_margin"]):
                    # print("obj too birght!!", flush=True)
                    # if obj.type != 'galaxy':
                    #     bright_obj += 1
                    #     obj.unload_SED()
                    #     continue
                    bright_obj += 1
                    obj.unload_SED()
                    continue
                if filt.is_too_dim(
                    mag=obj.getMagFilter(filt),
                    margin=self.config["obs_setting"]["mag_lim_margin"]):
                # if cut_filter.is_too_dim(mag=obj.param["mag_%s"%self.config["obs_setting"]["cut_in_band"].lower()]):
                    # print("obj too dim!!", flush=True)
                    dim_obj += 1
                    obj.unload_SED()
                    # print(obj.getMagFilter(filt))
                    continue

                if self.config["shear_setting"]["shear_type"] == "constant":
                    if obj.type == 'star':
                        obj.g1, obj.g2 = 0., 0.
                    else:
                        obj.g1, obj.g2 = self.g1_field, self.g2_field
                elif self.config["shear_setting"]["shear_type"] == "extra":
                    try:
                        # TODO: every object with individual shear from input catalog(s)
                        obj.g1, obj.g2 = self.g1_field[j], self.g2_field[j]
                    except:
                        # print("failed to load external shear.")
                        chip_output.logger.error("failed to load external shear.")
                        pass
                elif self.config["shear_setting"]["shear_type"] == "catalog":
                    pass
                else:
                    chip_output.logger.error("Unknown shear input")
                    raise ValueError("Unknown shear input")

                header_wcs = generateExtensionHeader(
                    xlen=chip.npix_x,
                    ylen=chip.npix_y,
                    ra=ra_cen,
                    dec=dec_cen,
                    pa=pointing.img_pa.deg,
                    gain=chip.gain,
                    readout=chip.read_noise,
                    dark=chip.dark_noise,
                    saturation=90000,
                    psize=chip.pix_scale,
                    row_num=chip.rowID,
                    col_num=chip.colID,
                    extName='raw')

                pos_img, offset, local_wcs, real_wcs = obj.getPosImg_Offset_WCS(img=chip.img, fdmodel=self.fd_model, chip=chip, verbose=False, img_header=header_wcs)
                # pos_img, offset, local_wcs, real_wcs = obj.getPosImg_Offset_WCS(img=chip.img, fdmodel=self.fd_model, chip=chip, verbose=True, img_header=header_wcs)
                if pos_img.x == -1 or pos_img.y == -1:
                    # Exclude object which is outside the chip area (after field distortion)
                    # print('obj_ra = ', obj.ra, 'obj_dec = ', obj.dec, 'obj_ra_orig = ', obj.ra_orig, 'obj_dec_orig = ',obj.dec_orig)
                    # print("obj missed: %s"%(obj.id))
                    chip_output.logger.error("obj missed: %s"%(obj.id))
                    missed_obj += 1
                    obj.unload_SED()
                    continue

                # Draw object & update output catalog
                try:
                    # chip_output.logger.info("current filter type: %s"%filt.filter_type)
                    if self.config["run_option"]["out_cat_only"]:
                        isUpdated = True
                        obj.real_pos = obj.getRealPos(chip.img, global_x=obj.posImg.x, global_y=obj.posImg.y,
                                        img_real_wcs=obj.real_wcs)
                        pos_shear = 0.
                    elif chip.survey_type == "photometric" and not self.config["run_option"]["out_cat_only"]:
                        isUpdated, pos_shear = obj.drawObj_multiband(
                            tel=self.tel,
                            pos_img=pos_img, 
                            psf_model=psf_model, 
                            bandpass_list=filt.bandpass_sub_list, 
                            filt=filt, 
                            chip=chip, 
                            g1=obj.g1, 
                            g2=obj.g2, 
                            exptime=pointing.exp_time
                            )

                    elif chip.survey_type == "spectroscopic" and not self.config["run_option"]["out_cat_only"]:
                        isUpdated, pos_shear = obj.drawObj_slitless(
                            tel=self.tel, 
                            pos_img=pos_img, 
                            psf_model=psf_model, 
                            bandpass_list=filt.bandpass_sub_list, 
                            filt=filt, 
                            chip=chip, 
                            g1=obj.g1, 
                            g2=obj.g2, 
                            exptime=pointing.exp_time,
                            normFilter=norm_filt)
                    if isUpdated:
                        # TODO: add up stats
                        # print("updating output catalog...")
                        chip_output.cat_add_obj(obj, pos_img, pos_shear)
                        pass
                    else:
                        # print("object omitted", flush=True)
                        continue
                except Exception as e:
                    # print(e)
                    traceback.print_exc()
                    chip_output.logger.error(e)
                    pass
                    
                # # [C6 TEST]
                # print("check running:1: pointing-{:} chip-{:} pid-{:} memory-{:6.2}GB".format(pointing.id, chip.chipID, os.getpid(), (psutil.Process(os.getpid()).memory_info().rss / 1024 / 1024 / 1024) ), flush=True)
                # print('draw object %s'%obj.id)
                # print('mag = %.3f'%obj.param['mag_use_normal'])
                # # Output memory usage
                # snapshot = tracemalloc.take_snapshot()
                # top_stats = snapshot.statistics('lineno')
                # for stat in top_stats[:10]:
                #     print(stat)

                # Unload SED:
                obj.unload_SED()
                del obj
                gc.collect()

            del psf_model
            del self.cat
            gc.collect()

        # print("check running:1: pointing-{:} chip-{:} pid-{:} memory-{:6.2}GB".format(pointing.id, chip.chipID, os.getpid(), (psutil.Process(os.getpid()).memory_info().rss / 1024 / 1024 / 1024) ), flush=True)
        chip_output.logger.info("check running:1: pointing-%d chip-%d pid-%d memory-%6.2fGB"%(pointing.id, chip.chipID, os.getpid(), (psutil.Process(os.getpid()).memory_info().rss / 1024 / 1024 / 1024) ))

        # Detector Effects
        # ===========================================================
        # whether to output zero, dark, flat calibration images.

        if not self.config["run_option"]["out_cat_only"]:
            chip.img = chip.addEffects(
                config=self.config, 
                img=chip.img, 
                chip_output=chip_output, 
                filt=filt, 
                ra_cen=pointing.ra, 
                dec_cen=pointing.dec,
                img_rot=pointing.img_pa,
                pointing_ID=pointing.id,
                timestamp_obs=pointing.timestamp,
                pointing_type=pointing.pointing_type,
                sky_map=sky_map, tel = self.tel,
                logger=chip_output.logger)
            
            if pointing.pointing_type == 'MS':
                datetime_obs = datetime.utcfromtimestamp(pointing.timestamp)
                date_obs = datetime_obs.strftime("%y%m%d")
                time_obs = datetime_obs.strftime("%H%M%S")
                h_prim = generatePrimaryHeader(
                    xlen=chip.npix_x, 
                    ylen=chip.npix_y, 
                    pointNum = str(pointing.id),
                    ra=pointing.ra, 
                    dec=pointing.dec, 
                    psize=chip.pix_scale, 
                    row_num=chip.rowID, 
                    col_num=chip.colID,
                    date=date_obs,
                    time_obs=time_obs,
                    exptime=pointing.exp_time,
                    im_type='SCI',
                    sat_pos=[pointing.sat_x, pointing.sat_y, pointing.sat_z],
                    sat_vel=[pointing.sat_vx, pointing.sat_vy, pointing.sat_vz])
                h_ext = generateExtensionHeader(
                    xlen=chip.npix_x, 
                    ylen=chip.npix_y, 
                    ra=pointing.ra, 
                    dec=pointing.dec, 
                    pa=pointing.img_pa.deg, 
                    gain=chip.gain, 
                    readout=chip.read_noise, 
                    dark=chip.dark_noise, 
                    saturation=90000, 
                    psize=chip.pix_scale, 
                    row_num=chip.rowID, 
                    col_num=chip.colID,
                    extName='raw')
                chip.img = galsim.Image(chip.img.array, dtype=np.uint16)
                hdu1 = fits.PrimaryHDU(header=h_prim)
                hdu1.add_checksum()
                hdu2 = fits.ImageHDU(chip.img.array, header=h_ext)
                hdu2.add_checksum()
                hdu1 = fits.HDUList([hdu1, hdu2])
                fname = os.path.join(chip_output.subdir, h_prim['FILENAME'] + '.fits')
                hdu1.writeto(fname, output_verify='ignore', overwrite=True)
                # print("# objects that are too bright %d out of %d"%(bright_obj, self.nobj))
                # print("# objects that are too dim %d out of %d"%(dim_obj, self.nobj))
                # print("# objects that are missed %d out of %d"%(missed_obj, self.nobj))
                chip_output.logger.info("# objects that are too bright %d out of %d"%(bright_obj, self.nobj))
                chip_output.logger.info("# objects that are too dim %d out of %d"%(dim_obj, self.nobj))
                chip_output.logger.info("# objects that are missed %d out of %d"%(missed_obj, self.nobj))
        del chip.img

        # print("check running:2: pointing-{:} chip-{:} pid-{:} memory-{:6.2}GB".format(pointing.id, chip.chipID, os.getpid(), (psutil.Process(os.getpid()).memory_info().rss / 1024 / 1024 / 1024) ), flush=True)
        chip_output.logger.info("check running:2: pointing-%d chip-%d pid-%d memory-%6.2fGB"%(pointing.id, chip.chipID, os.getpid(), (psutil.Process(os.getpid()).memory_info().rss / 1024 / 1024 / 1024) ))

    def runExposure_MPI_PointingList(self, pointing_list, shear_cat_file=None, chips=None, use_mpi=False):
        if use_mpi:
            comm = MPI.COMM_WORLD
            ind_thread = comm.Get_rank()
            num_thread = comm.Get_size()

        if chips is None:
            nchips_per_fp = len(self.chip_list)
            run_chips = self.chip_list
            run_filts = self.filter_list
        else:
            # Only run a particular set of chips
            run_chips = []
            run_filts = []
            nchips_per_fp = len(chips)
            for ichip in range(len(self.chip_list)):
                chip = self.chip_list[ichip]
                filt = self.filter_list[ichip]
                if chip.chipID in chips:
                    run_chips.append(chip)
                    run_filts.append(filt)

        for ipoint in range(len(pointing_list)):
            for ichip in range(nchips_per_fp):
                i = ipoint*nchips_per_fp + ichip
                pointing = pointing_list[ipoint]
                pointing_ID = pointing.id
                if use_mpi:
                    if i % num_thread != ind_thread:
                        continue

                pid = os.getpid()

                sub_img_dir, prefix = makeSubDir_PointingList(path_dict=self.path_dict, config=self.config, pointing_ID=pointing_ID)

                chip = run_chips[ichip]
                filt = run_filts[ichip]
                # print("running pointing#%d, chip#%d, at PID#%d..."%(pointing_ID, chip.chipID, pid), flush=True)
                chip_output = ChipOutput(
                    config=self.config, 
                    focal_plane=self.focal_plane, 
                    chip=chip, 
                    filt=filt,  
                    exptime=pointing.exp_time,
                    pointing_type=pointing.pointing_type,
                    pointing_ID=pointing_ID,  
                    subdir=sub_img_dir,
                    prefix=prefix)
                chip_output.logger.info("running pointing#%d, chip#%d, at PID#%d..."%(pointing_ID, chip.chipID, pid))
                self.run_one_chip(
                    chip=chip, 
                    filt=filt, 
                    chip_output=chip_output, 
                    pointing=pointing,
                    cat_dir=self.path_dict["cat_dir"])
                print("finished running chip#%d..."%(chip.chipID), flush=True)
                chip_output.logger.info("finished running chip#%d..."%(chip.chipID))
                for handler in chip_output.logger.handlers[:]:
                    chip_output.logger.removeHandler(handler)
                gc.collect()