def prepare_timeseries(outfile,
unw_file,
metadata,
processor,
baseline_dir=None,
box=None):
print('-' * 50)
print('preparing timeseries file: {}'.format(outfile))
# copy metadata to meta
meta = {key: value for key, value in metadata.items()}
phase2range = float(meta['WAVELENGTH']) / (4. * np.pi)
# grab date list from the filename
unw_files = sorted(glob.glob(unw_file))
date12_list = [os.path.splitext(os.path.basename(i))[0] for i in unw_files]
num_file = len(unw_files)
print('number of unwrapped interferograms: {}'.format(num_file))
ref_date = date12_list[0].split('_')[0]
date_list = [ref_date] + [date12.split('_')[1] for date12 in date12_list]
num_date = len(date_list)
print('number of acquisitions: {}\n{}'.format(num_date, date_list))
# baseline info
if baseline_dir is not None:
# read baseline data
baseline_dict = isce_utils.read_baseline_timeseries(
baseline_dir, processor=processor, ref_date=ref_date)
# dict to array
pbase = np.zeros(num_date, dtype=np.float32)
for i in range(num_date):
pbase_top, pbase_bottom = baseline_dict[date_list[i]]
pbase[i] = (pbase_top + pbase_bottom) / 2.0
# size info
if not box:
box = (0, 0, int(meta['WIDTH']), int(meta['LENGTH']))
kwargs = dict(xoff=box[0],
yoff=box[1],
win_xsize=box[2] - box[0],
win_ysize=box[3] - box[1])
# define dataset structure
dates = np.array(date_list, dtype=np.string_)
ds_name_dict = {
"date": [dates.dtype, (num_date, ), dates],
"bperp": [np.float32, (num_date, ), pbase],
"timeseries":
[np.float32, (num_date, box[3] - box[1], box[2] - box[0]), None],
}
# initiate HDF5 file
meta["FILE_TYPE"] = "timeseries"
meta["UNIT"] = "m"
meta['REF_DATE'] = ref_date
writefile.layout_hdf5(outfile, ds_name_dict, metadata=meta)
# writing data to HDF5 file
print('writing data to HDF5 file {} with a mode ...'.format(outfile))
with h5py.File(outfile, "a") as f:
prog_bar = ptime.progressBar(maxValue=num_file)
for i, unw_file in enumerate(unw_files):
# read data using gdal
ds = gdal.Open(unw_file, gdal.GA_ReadOnly)
data = np.array(ds.GetRasterBand(2).ReadAsArray(**kwargs),
dtype=np.float32)
f["timeseries"][i + 1] = data * phase2range
prog_bar.update(i + 1, suffix=date12_list[i])
prog_bar.close()
print('set value at the first acquisition to ZERO.')
f["timeseries"][0] = 0.
print('finished writing to HDF5 file: {}'.format(outfile))
return outfile
def main(iargs=None):
inps = cmd_line_parse(iargs)
if inps.updateMode:
print('update mode: ON')
else:
print('update mode: OFF')
# extract metadata
meta = extract_metadata(inps.unwFile)
box, meta = read_subset_box(inps.template_file, meta)
length = int(meta["LENGTH"])
width = int(meta["WIDTH"])
num_pair = meta["NUMBER_OF_PAIRS"]
# prepare output directory
out_dir = os.path.dirname(inps.outfile[0])
os.makedirs(out_dir, exist_ok=True)
########## output file 1 - ifgramStack
# define dataset structure for ifgramStack
dsNameDict = {
"date": (np.dtype('S8'), (num_pair, 2)),
"dropIfgram": (np.bool_, (num_pair, )),
"bperp": (np.float32, (num_pair, )),
"unwrapPhase": (np.float32, (num_pair, length, width)),
"coherence": (np.float32, (num_pair, length, width)),
"connectComponent": (np.int16, (num_pair, length, width)),
}
if run_or_skip(inps, dsNameDict, out_file=inps.outfile[0]) == 'run':
# initiate h5 file with defined structure
meta['FILE_TYPE'] = 'ifgramStack'
writefile.layout_hdf5(inps.outfile[0],
dsNameDict,
meta,
compression=inps.compression)
# write data to h5 file in disk
write_ifgram_stack(inps.outfile[0],
inps.unwFile,
inps.corFile,
inps.connCompFile,
box=box)
########## output file 2 - geometryGeo
# define dataset structure for geometry
dsNameDict = {
"height": (np.float32, (length, width)),
"incidenceAngle": (np.float32, (length, width)),
"slantRangeDistance": (np.float32, (length, width)),
}
if inps.azAngleFile is not None:
dsNameDict["azimuthAngle"] = (np.float32, (length, width))
if inps.waterMaskFile is not None:
dsNameDict["waterMask"] = (np.bool_, (length, width))
if run_or_skip(inps, dsNameDict, out_file=inps.outfile[1]) == 'run':
# initiate h5 file with defined structure
meta['FILE_TYPE'] = 'geometry'
writefile.layout_hdf5(inps.outfile[1],
dsNameDict,
meta,
compression=inps.compression)
# write data to disk
write_geometry(inps.outfile[1],
demFile=inps.demFile,
incAngleFile=inps.incAngleFile,
azAngleFile=inps.azAngleFile,
waterMaskFile=inps.waterMaskFile,
box=box)
print('-' * 50)
return inps.outfile
def ifgram_inversion(inps=None):
"""Phase triangulatino of small baseline interferograms
Parameters: inps - namespace
Example: inps = cmd_line_parse()
ifgram_inversion(inps)
"""
if not inps:
inps = cmd_line_parse()
start_time = time.time()
## 1. input info
stack_obj = ifgramStack(inps.ifgramStackFile)
stack_obj.open(print_msg=False)
date12_list = stack_obj.get_date12_list(dropIfgram=True)
date_list = stack_obj.get_date_list(dropIfgram=True)
length, width = stack_obj.length, stack_obj.width
# 1.1 read values on the reference pixel
inps.refPhase = stack_obj.get_reference_phase(
unwDatasetName=inps.obsDatasetName,
skip_reference=inps.skip_ref,
dropIfgram=True)
# 1.2 design matrix
A = stack_obj.get_design_matrix4timeseries(date12_list)[0]
num_ifgram, num_date = A.shape[0], A.shape[1] + 1
inps.numIfgram = num_ifgram
# 1.3 print key setup info
msg = '-------------------------------------------------------------------------------\n'
if inps.minNormVelocity:
suffix = 'deformation velocity'
else:
suffix = 'deformation phase'
msg += 'least-squares solution with L2 min-norm on: {}\n'.format(suffix)
msg += 'minimum redundancy: {}\n'.format(inps.minRedundancy)
msg += 'weight function: {}\n'.format(inps.weightFunc)
if inps.maskDataset:
if inps.maskDataset in ['coherence', 'offsetSNR']:
suffix = '{} < {}'.format(inps.maskDataset, inps.maskThreshold)
else:
suffix = '{} == 0'.format(inps.maskDataset)
msg += 'mask out pixels with: {}\n'.format(suffix)
else:
msg += 'mask: no\n'
if np.linalg.matrix_rank(A) < A.shape[1]:
msg += '***WARNING: the network is NOT fully connected.\n'
msg += '\tInversion result can be biased!\n'
msg += '\tContinue to use SVD to resolve the offset between different subsets.\n'
msg += '-------------------------------------------------------------------------------'
print(msg)
print('number of interferograms: {}'.format(num_ifgram))
print('number of acquisitions : {}'.format(num_date))
print('number of lines : {}'.format(length))
print('number of columns : {}'.format(width))
## 2. prepare output
# 2.1 metadata
meta = dict(stack_obj.metadata)
for key in configKeys:
meta[key_prefix + key] = str(vars(inps)[key])
# 2.2 instantiate time-series
dsNameDict = {
"date": (np.dtype('S8'), (num_date, )),
"bperp": (np.float32, (num_date, )),
"timeseries": (np.float32, (num_date, length, width)),
}
meta['FILE_TYPE'] = 'timeseries'
meta['UNIT'] = 'm'
meta['REF_DATE'] = date_list[0]
ts_obj = timeseries(inps.tsFile)
ts_obj.layout_hdf5(dsNameDict, meta)
# write date time-series
date_list_utf8 = [dt.encode('utf-8') for dt in date_list]
writefile.write_hdf5_block(inps.tsFile, date_list_utf8, datasetName='date')
# write bperp time-series
pbase = stack_obj.get_perp_baseline_timeseries(dropIfgram=True)
writefile.write_hdf5_block(inps.tsFile, pbase, datasetName='bperp')
# 2.3 instantiate temporal coherence
dsNameDict = {"temporalCoherence": (np.float32, (length, width))}
meta['FILE_TYPE'] = 'temporalCoherence'
meta['UNIT'] = '1'
meta.pop('REF_DATE')
writefile.layout_hdf5(inps.tempCohFile, dsNameDict, metadata=meta)
# 2.4 instantiate number of inverted observations
dsNameDict = {"mask": (np.float32, (length, width))}
meta['FILE_TYPE'] = 'mask'
meta['UNIT'] = '1'
writefile.layout_hdf5(inps.numInvFile, dsNameDict, metadata=meta)
## 3. run the inversion / estimation and write to disk
# 3.1 split ifgram_file into blocks to save memory
box_list, num_box = split2boxes(inps.ifgramStackFile,
memory_size=inps.memorySize)
# 3.2 prepare the input arguments for *_patch()
data_kwargs = {
"ifgram_file": inps.ifgramStackFile,
"ref_phase": inps.refPhase,
"obs_ds_name": inps.obsDatasetName,
"weight_func": inps.weightFunc,
"min_norm_velocity": inps.minNormVelocity,
"water_mask_file": inps.waterMaskFile,
"mask_ds_name": inps.maskDataset,
"mask_threshold": inps.maskThreshold,
"min_redundancy": inps.minRedundancy
}
# 3.3 invert / write block-by-block
for i, box in enumerate(box_list):
box_width = box[2] - box[0]
box_length = box[3] - box[1]
if num_box > 1:
print('\n------- processing patch {} out of {} --------------'.
format(i + 1, num_box))
print('box width: {}'.format(box_width))
print('box length: {}'.format(box_length))
# update box argument in the input data
data_kwargs['box'] = box
if inps.cluster == 'no':
# non-parallel
ts, temp_coh, num_inv_ifg = ifgram_inversion_patch(
**data_kwargs)[:-1]
else:
# parallel
print('\n\n------- start parallel processing using Dask -------')
# initiate the output data
ts = np.zeros((num_date, box_length, box_width), np.float32)
temp_coh = np.zeros((box_length, box_width), np.float32)
num_inv_ifg = np.zeros((box_length, box_width), np.float32)
# initiate dask cluster and client
cluster_obj = cluster.DaskCluster(inps.cluster,
inps.numWorker,
config_name=inps.config)
cluster_obj.open()
# run dask
ts, temp_coh, num_inv_ifg = cluster_obj.run(
func=ifgram_inversion_patch,
func_data=data_kwargs,
results=[ts, temp_coh, num_inv_ifg])
# close dask cluster and client
cluster_obj.close()
print('------- finished parallel processing -------\n\n')
# write the block to disk
# with 3D block in [z0, z1, y0, y1, x0, x1]
# and 2D block in [y0, y1, x0, x1]
# time-series - 3D
block = [0, num_date, box[1], box[3], box[0], box[2]]
writefile.write_hdf5_block(inps.tsFile,
data=ts,
datasetName='timeseries',
block=block)
# temporal coherence - 2D
block = [box[1], box[3], box[0], box[2]]
writefile.write_hdf5_block(inps.tempCohFile,
data=temp_coh,
datasetName='temporalCoherence',
block=block)
# number of inverted obs - 2D
writefile.write_hdf5_block(inps.numInvFile,
data=num_inv_ifg,
datasetName='mask',
block=block)
m, s = divmod(time.time() - start_time, 60)
print('time used: {:02.0f} mins {:02.1f} secs.\n'.format(m, s))
# 3.4 update output data on the reference pixel
if not inps.skip_ref:
# grab ref_y/x
ref_y = int(stack_obj.metadata['REF_Y'])
ref_x = int(stack_obj.metadata['REF_X'])
print('-' * 50)
print('update values on the reference pixel: ({}, {})'.format(
ref_y, ref_x))
print('set temporal coherence on the reference pixel to 1.')
with h5py.File(inps.tempCohFile, 'r+') as f:
f['temporalCoherence'][ref_y, ref_x] = 1.
print('set # of observations on the reference pixel as {}'.format(
num_ifgram))
with h5py.File(inps.numInvFile, 'r+') as f:
f['mask'][ref_y, ref_x] = num_ifgram
m, s = divmod(time.time() - start_time, 60)
print('time used: {:02.0f} mins {:02.1f} secs.\n'.format(m, s))
return
def main(iargs=None):
"""
Overwrite filtered SLC images in Isce merged/SLC directory.
"""
Parser = MinoPyParser(iargs, script='generate_temporal_coherence')
inps = Parser.parse()
dateStr = datetime.datetime.strftime(datetime.datetime.now(),
'%Y%m%d:%H%M%S')
if not iargs is None:
msg = os.path.basename(__file__) + ' ' + ' '.join(iargs[:])
string = dateStr + " * " + msg
print(string)
else:
msg = os.path.basename(__file__) + ' ' + ' '.join(sys.argv[1::])
string = dateStr + " * " + msg
print(string)
start_time = time.time()
os.chdir(inps.work_dir)
minopy_dir = os.path.dirname(inps.work_dir)
minopy_template_file = os.path.join(minopy_dir, 'minopyApp.cfg')
inps.ifgramStackFile = os.path.join(inps.work_dir, 'inputs/ifgramStack.h5')
template = readfile.read_template(minopy_template_file)
if template['minopy.timeseries.tempCohType'] == 'auto':
template['minopy.timeseries.tempCohType'] = 'full'
atr = {}
atr['minopy.timeseries.tempCohType'] = template[
'minopy.timeseries.tempCohType']
ut.add_attribute(inps.ifgramStackFile, atr)
# check if input observation dataset exists.
stack_obj = ifgramStack(inps.ifgramStackFile)
stack_obj.open(print_msg=False)
metadata = stack_obj.get_metadata()
length, width = stack_obj.length, stack_obj.width
inps.invQualityFile = 'temporalCoherence.h5'
mintpy_mask_file = os.path.join(inps.work_dir, 'maskTempCoh.h5')
quality_name = os.path.join(
minopy_dir, 'inverted/tempCoh_{}'.format(
template['minopy.timeseries.tempCohType']))
quality = np.memmap(quality_name,
mode='r',
dtype='float32',
shape=(length, width))
# inps.waterMaskFile = os.path.join(minopy_dir, 'waterMask.h5')
inps.waterMaskFile = None
water_mask = np.ones(quality.shape, dtype=np.int8)
if template['minopy.timeseries.waterMask'] != 'auto':
inps.waterMaskFile = template['minopy.timeseries.waterMask']
if os.path.exists(inps.waterMaskFile):
with h5py.File(inps.waterMaskFile, 'r') as f2:
if 'waterMask' in f2:
water_mask = f2['waterMask'][:, :]
else:
water_mask = f2['mask'][:, :]
if inps.shadow_mask:
if os.path.exists(os.path.join(minopy_dir, 'shadow_mask.h5')):
with h5py.File(os.path.join(minopy_dir, 'shadow_mask.h5'),
'r') as f2:
shadow_mask = f2['mask'][:, :]
water_mask = water_mask * shadow_mask
inv_quality = np.zeros((quality.shape[0], quality.shape[1]))
inv_quality_name = 'temporalCoherence'
inv_quality[:, :] = quality[:, :]
inv_quality[inv_quality <= 0] = np.nan
inv_quality[water_mask < 0.5] = np.nan
if os.path.exists(mintpy_mask_file):
mintpy_mask = readfile.read(mintpy_mask_file, datasetName='mask')[0]
inv_quality[mintpy_mask == 0] = np.nan
if not os.path.exists(inps.invQualityFile):
metadata['UNIT'] = '1'
metadata['FILE_TYPE'] = inv_quality_name
if 'REF_DATE' in metadata:
metadata.pop('REF_DATE')
ds_name_dict = {metadata['FILE_TYPE']: [np.float32, (length, width)]}
writefile.layout_hdf5(inps.invQualityFile,
ds_name_dict,
metadata=metadata)
# write the block to disk
# with 3D block in [z0, z1, y0, y1, x0, x1]
# and 2D block in [y0, y1, x0, x1]
block = [0, length, 0, width]
writefile.write_hdf5_block(inps.invQualityFile,
data=inv_quality,
datasetName=inv_quality_name,
block=block)
get_phase_linking_coherence_mask(metadata, inps.work_dir)
m, s = divmod(time.time() - start_time, 60)
print('time used: {:02.0f} mins {:02.1f} secs.\n'.format(m, s))
return
def run_geocode(inps):
"""geocode all input files"""
start_time = time.time()
# feed the largest file for resample object initiation
ind_max = np.argmax([os.path.getsize(i) for i in inps.file])
# prepare geometry for geocoding
res_obj = resample(lut_file=inps.lookupFile,
src_file=inps.file[ind_max],
SNWE=inps.SNWE,
lalo_step=inps.laloStep,
interp_method=inps.interpMethod,
fill_value=inps.fillValue,
nprocs=inps.nprocs,
max_memory=inps.maxMemory,
software=inps.software,
print_msg=True)
res_obj.open()
res_obj.prepare()
# resample input files one by one
for infile in inps.file:
print('-' * 50 + '\nresampling file: {}'.format(infile))
ext = os.path.splitext(infile)[1]
atr = readfile.read_attribute(infile, datasetName=inps.dset)
outfile = auto_output_filename(infile, inps)
# update_mode
if inps.updateMode:
print('update mode: ON')
if ut.run_or_skip(outfile, in_file=[infile,
inps.lookupFile]) == 'skip':
continue
## prepare output
# update metadata
if inps.radar2geo:
atr = attr.update_attribute4radar2geo(atr, res_obj=res_obj)
else:
atr = attr.update_attribute4geo2radar(atr, res_obj=res_obj)
# instantiate output file
file_is_hdf5 = os.path.splitext(infile)[1] in ['.h5', '.he5']
if file_is_hdf5:
writefile.layout_hdf5(outfile, metadata=atr, ref_file=infile)
else:
dsDict = dict()
## run
dsNames = readfile.get_dataset_list(infile, datasetName=inps.dset)
maxDigit = max([len(i) for i in dsNames])
for dsName in dsNames:
if not file_is_hdf5:
dsDict[dsName] = np.zeros((res_obj.length, res_obj.width))
# loop for block-by-block IO
for i in range(res_obj.num_box):
src_box = res_obj.src_box_list[i]
dest_box = res_obj.dest_box_list[i]
# read
print('-' * 50 +
'\nreading {d:<{w}} in block {b} from {f} ...'.format(
d=dsName,
w=maxDigit,
b=src_box,
f=os.path.basename(infile)))
data = readfile.read(infile,
datasetName=dsName,
box=src_box,
print_msg=False)[0]
# resample
data = res_obj.run_resample(src_data=data, box_ind=i)
# write / save block data
if data.ndim == 3:
block = [
0, data.shape[0], dest_box[1], dest_box[3],
dest_box[0], dest_box[2]
]
else:
block = [
dest_box[1], dest_box[3], dest_box[0], dest_box[2]
]
if file_is_hdf5:
print('write data in block {} to file: {}'.format(
block, outfile))
writefile.write_hdf5_block(outfile,
data=data,
datasetName=dsName,
block=block,
print_msg=False)
else:
dsDict[dsName][block[0]:block[1], block[2]:block[3]] = data
# for binary file: ensure same data type
if not file_is_hdf5:
dsDict[dsName] = np.array(dsDict[dsName], dtype=data.dtype)
# write binary file
if not file_is_hdf5:
writefile.write(dsDict,
out_file=outfile,
metadata=atr,
ref_file=infile)
m, s = divmod(time.time() - start_time, 60)
print('time used: {:02.0f} mins {:02.1f} secs.\n'.format(m, s))
return outfile
def prepare_timeseries(outfile, unw_file, metadata, processor, baseline_dir=None):
print('-'*50)
print('preparing timeseries file: {}'.format(outfile))
# copy metadata to meta
meta = {key : value for key, value in metadata.items()}
phase2range = -1. * float(meta['WAVELENGTH']) / (4. * np.pi)
# grab date list from the filename
unw_files = sorted(glob.glob(unw_file))
date12_list = [os.path.splitext(os.path.basename(i))[0] for i in unw_files]
num_file = len(unw_files)
print('number of unwrapped interferograms: {}'.format(num_file))
ref_date = date12_list[0].split('_')[0]
date_list = [ref_date] + [date12.split('_')[1] for date12 in date12_list]
num_date = len(date_list)
print('number of acquisitions: {}\n{}'.format(num_date, date_list))
# define dataset structure
length, width = int(meta['LENGTH']), int(meta['WIDTH'])
dsNameDict = {
"date" : (np.dtype("S8"), (num_date,)),
"timeseries" : (np.float32, (num_date, length, width))
}
# baseline info
baseline_dict = {}
if baseline_dir is not None:
# read baseline data
baseline_dict = isce_utils.read_baseline_timeseries(baseline_dir,
processor=processor,
ref_date=ref_date)
# dict to array
pbase = np.zeros(num_date, dtype=np.float32)
for i in range(num_date):
pbase_top, pbase_bottom = baseline_dict[date_list[i]]
pbase[i] = (pbase_top + pbase_bottom) / 2.0
# update dataset structure
dsNameDict["bperp"] = (np.float32, (num_date,))
# initiate HDF5 file
meta["FILE_TYPE"] = "timeseries"
meta["UNIT"] = "m"
meta['REF_DATE'] = ref_date
writefile.layout_hdf5(outfile, dsNameDict, meta)
# writing data to HDF5 file
print('writing data to HDF5 file {} with a mode ...'.format(outfile))
with h5py.File(outfile, "a") as f:
f["date"][:,] = np.array([np.string_(i) for i in date_list])
f["bperp"][:,] = pbase
prog_bar = ptime.progressBar(maxValue=num_file)
for i in range(num_file):
# read data using gdal
ds = gdal.Open(unw_files[i], gdal.GA_ReadOnly)
data = np.array(ds.GetRasterBand(2).ReadAsArray())
f["timeseries"][i+1] = data * phase2range
prog_bar.update(i+1, suffix=date12_list[i])
prog_bar.close()
print('set value at the first acquisition to ZERO.')
f["timeseries"][0] = 0.
print('finished writing to HDF5 file: {}'.format(outfile))
return outfile
def main(iargs=None):
inps = cmd_line_parse(iargs)
start_time = time.time()
if inps.updateMode:
print('update mode: ON')
else:
print('update mode: OFF')
# extract metadata
meta = extract_metadata(inps.unwFile)
box, meta = read_subset_box(inps.template_file, meta)
if inps.xstep * inps.ystep > 1:
meta = attr.update_attribute4multilook(meta,
lks_y=inps.ystep,
lks_x=inps.xstep)
length = int(meta["LENGTH"])
width = int(meta["WIDTH"])
num_pair = int(meta["NUMBER_OF_PAIRS"])
# prepare output directory
out_dir = os.path.dirname(inps.outfile[0])
os.makedirs(out_dir, exist_ok=True)
########## output file 1 - ifgramStack
# define dataset structure for ifgramStack
dsNameDict = {
"date": (np.dtype('S8'), (num_pair, 2)),
"dropIfgram": (np.bool_, (num_pair, )),
"bperp": (np.float32, (num_pair, )),
"unwrapPhase": (np.float32, (num_pair, length, width)),
"coherence": (np.float32, (num_pair, length, width)),
"connectComponent": (np.int16, (num_pair, length, width)),
}
if inps.magFile is not None:
dsNameDict['magnitude'] = (np.float32, (num_pair, length, width))
if run_or_skip(inps, dsNameDict, out_file=inps.outfile[0]) == 'run':
# initiate h5 file with defined structure
meta['FILE_TYPE'] = 'ifgramStack'
writefile.layout_hdf5(inps.outfile[0],
dsNameDict,
meta,
compression=inps.compression)
# write data to h5 file in disk
write_ifgram_stack(inps.outfile[0],
unwStack=inps.unwFile,
cohStack=inps.corFile,
connCompStack=inps.connCompFile,
ampStack=inps.magFile,
box=box,
xstep=inps.xstep,
ystep=inps.ystep)
########## output file 2 - geometryGeo
# define dataset structure for geometry
dsNameDict = {
"height": (np.float32, (length, width)),
"incidenceAngle": (np.float32, (length, width)),
"slantRangeDistance": (np.float32, (length, width)),
}
if inps.azAngleFile is not None:
dsNameDict["azimuthAngle"] = (np.float32, (length, width))
if inps.waterMaskFile is not None:
dsNameDict["waterMask"] = (np.bool_, (length, width))
if run_or_skip(inps, dsNameDict, out_file=inps.outfile[1]) == 'run':
# initiate h5 file with defined structure
meta['FILE_TYPE'] = 'geometry'
writefile.layout_hdf5(inps.outfile[1],
dsNameDict,
meta,
compression=inps.compression)
# write data to disk
write_geometry(inps.outfile[1],
demFile=inps.demFile,
incAngleFile=inps.incAngleFile,
azAngleFile=inps.azAngleFile,
waterMaskFile=inps.waterMaskFile,
box=box,
xstep=inps.xstep,
ystep=inps.ystep)
print('-' * 50)
# time info
m, s = divmod(time.time() - start_time, 60)
print('time used: {:02.0f} mins {:02.1f} secs.'.format(m, s))
return inps.outfile
def calculate_delay_timeseries(tropo_file, dis_file, geom_file, GACOS_dir):
"""calculate delay time-series and write to HDF5 file"""
## get list of dates
atr = readfile.read_attribute(dis_file)
ftype = atr['FILE_TYPE']
if ftype == 'timeseries':
date_list = timeseries(dis_file).get_date_list()
elif ftype == '.unw':
date12 = readfile.read_attribute(dis_file)['DATE12']
date_list = ptime.yyyymmdd(date12.split('-'))
else:
raise ValueError(
'un-supported displacement file type: {}'.format(ftype))
# list of dates --> list of ztd files
ztd_files = [
os.path.join(GACOS_dir, '{}.ztd'.format(i)) for i in date_list
]
# check missing ztd files
flag = np.ones(len(date_list), dtype=np.bool_)
for i in range(len(date_list)):
if not os.path.isfile(ztd_files[i]):
print('WARNING: {} file not found, ignore it and continue'.format(
ztd_files[i]))
flag[i] = False
if np.any(flag == 0):
date_list = np.array(date_list)[flag].tolist()
ztd_files = np.array(ztd_files)[flag].tolist()
## update_mode
def get_dataset_size(fname):
atr = readfile.read_attribute(fname)
return (atr['LENGTH'], atr['WIDTH'])
def run_or_skip(ztd_files, tropo_file, geom_file):
print('update mode: ON')
print('output file: {}'.format(tropo_file))
flag = 'skip'
# check existance and modification time
if ut.run_or_skip(out_file=tropo_file,
in_file=ztd_files,
print_msg=False) == 'run':
flag = 'run'
print(
'1) output file either do NOT exist or is NOT newer than all ZTD files.'
)
else:
print('1) output file exists and is newer than all ZTD files.')
# check dataset size in space / time
date_list = [str(re.findall('\d{8}', i)[0]) for i in ztd_files]
if (get_dataset_size(tropo_file) != get_dataset_size(geom_file)
or any(i not in timeseries(tropo_file).get_date_list()
for i in date_list)):
flag = 'run'
print(
'2) output file does NOT have the same len/wid as the geometry file {} or does NOT contain all dates'
.format(geom_file))
else:
print(
'2) output file has the same len/wid as the geometry file and contains all dates'
)
# check if output file is fully written
with h5py.File(tropo_file, 'r') as f:
if np.all(f['timeseries'][-1, :, :] == 0):
flag = 'run'
print('3) output file is NOT fully written.')
else:
print('3) output file is fully written.')
# result
print('run or skip: {}'.format(flag))
return flag
if run_or_skip(ztd_files, tropo_file, geom_file) == 'skip':
return
## prepare output file
# metadata
atr['FILE_TYPE'] = 'timeseries'
atr['UNIT'] = 'm'
# remove metadata related with double reference
# because absolute delay is calculated and saved
for key in ['REF_DATE', 'REF_X', 'REF_Y', 'REF_LAT', 'REF_LON']:
if key in atr.keys():
atr.pop(key)
# instantiate time-series
length, width = int(atr['LENGTH']), int(atr['WIDTH'])
num_date = len(date_list)
dates = np.array(date_list, dtype=np.string_)
ds_name_dict = {
"date": [dates.dtype, (num_date, ), dates],
"timeseries": [np.float32, (num_date, length, width), None],
}
writefile.layout_hdf5(tropo_file, ds_name_dict, metadata=atr)
## calculate phase delay
# read geometry
print('read incidenceAngle from file: {}'.format(geom_file))
inc_angle = readfile.read(geom_file, datasetName='incidenceAngle')[0]
cos_inc_angle = np.cos(inc_angle * np.pi / 180.0)
if 'Y_FIRST' in atr.keys():
pts_new = None
else:
# pixel coordinates in geometry file
print('get pixel coordinates in geometry file')
lats, lons = ut.get_lat_lon(atr, geom_file)
pts_new = np.hstack((lats.reshape(-1, 1), lons.reshape(-1, 1)))
# loop for date-by-date IO
prog_bar = ptime.progressBar(maxValue=num_date)
for i in range(num_date):
date_str = date_list[i]
ztd_file = ztd_files[i]
# calc delay
if 'Y_FIRST' in atr.keys():
delay = get_delay_geo(ztd_file, atr, cos_inc_angle)
else:
delay = get_delay_radar(ztd_file, cos_inc_angle, pts_new)
# write delay to file
block = [i, i + 1, 0, length, 0, width]
writefile.write_hdf5_block(tropo_file,
data=delay,
datasetName='timeseries',
block=block,
print_msg=False)
prog_bar.update(i + 1, suffix=os.path.basename(ztd_file))
prog_bar.close()
return tropo_file
def multilook_file(infile,
lks_y,
lks_x,
outfile=None,
method='average',
margin=[0, 0, 0, 0]):
""" Multilook input file
Parameters: infile - str, path of input file to be multilooked.
lks_y - int, number of looks in y / row direction.
lks_x - int, number of looks in x / column direction.
margin - list of 4 int, number of pixels to be skipped during multilooking.
useful for offset product, where the marginal pixels are ignored during
cross correlation matching.
outfile - str, path of output file
Returns: outfile - str, path of output file
"""
lks_y = int(lks_y)
lks_x = int(lks_x)
# input file info
atr = readfile.read_attribute(infile)
length, width = int(atr['LENGTH']), int(atr['WIDTH'])
k = atr['FILE_TYPE']
print('multilooking {} {} file: {}'.format(atr['PROCESSOR'], k, infile))
print('number of looks in y / azimuth direction: %d' % lks_y)
print('number of looks in x / range direction: %d' % lks_x)
print('multilook method: {}'.format(method))
# margin --> box
if margin is not [0, 0, 0, 0]: # top, bottom, left, right
box = (margin[2], margin[0], width - margin[3], length - margin[1])
print(
'number of pixels to skip in top/bottom/left/right boundaries: {}'.
format(margin))
else:
box = (0, 0, width, length)
# output file name
ext = os.path.splitext(infile)[1]
if not outfile:
if os.getcwd() == os.path.dirname(os.path.abspath(infile)):
outfile = os.path.splitext(infile)[0] + '_' + str(
lks_y) + 'alks_' + str(lks_x) + 'rlks' + ext
else:
outfile = os.path.basename(infile)
# update metadata
atr = multilook_attribute(atr, lks_y, lks_x, box=box)
if ext in ['.h5', '.he5']:
writefile.layout_hdf5(outfile, metadata=atr, ref_file=infile)
# read source data and multilooking
dsNames = readfile.get_dataset_list(infile)
maxDigit = max([len(i) for i in dsNames])
dsDict = dict()
for dsName in dsNames:
print('multilooking {d:<{w}} from {f} ...'.format(
d=dsName, w=maxDigit, f=os.path.basename(infile)))
# split in Y/row direction for IO for HDF5 only
if ext in ['.h5', '.he5']:
row_step = 200
else:
row_step = box[3] - box[1]
num_step = int(np.ceil((box[3] - box[1]) / (row_step * lks_y)))
for i in range(num_step):
r0 = box[1] + row_step * lks_y * i
r1 = box[1] + row_step * lks_y * (i + 1)
r1 = min(r1, box[3])
# IO box
box_i = (box[0], r0, box[2], r1)
box_o = (int((box[0] - box[0]) / lks_x), int(
(r0 - box[1]) / lks_y), int(
(box[2] - box[0]) / lks_x), int((r1 - box[1]) / lks_y))
print('box: {}'.format(box_o))
# read / multilook
if method == 'nearest':
data = readfile.read(infile,
datasetName=dsName,
box=box_i,
xstep=lks_x,
ystep=lks_y,
print_msg=False)[0]
# fix the size discrepency between average / nearest method
out_len = box_o[3] - box_o[1]
out_wid = box_o[2] - box_o[0]
if data.ndim == 3:
data = data[:, :out_len, :out_wid]
else:
data = data[:out_len, :out_wid]
else:
data = readfile.read(infile,
datasetName=dsName,
box=box_i,
print_msg=False)[0]
# keep timeseries data as 3D matrix when there is only one acquisition
# because readfile.read() will squeeze it to 2D
if atr['FILE_TYPE'] == 'timeseries' and len(data.shape) == 2:
data = np.reshape(data, (1, data.shape[0], data.shape[1]))
data = multilook_data(data, lks_y, lks_x)
# output block
if data.ndim == 3:
block = [
0, data.shape[0], box_o[1], box_o[3], box_o[0], box_o[2]
]
else:
block = [box_o[1], box_o[3], box_o[0], box_o[2]]
# write
if ext in ['.h5', '.he5']:
writefile.write_hdf5_block(outfile,
data=data,
datasetName=dsName,
block=block,
print_msg=False)
else:
dsDict[dsName] = data
# for binary file with 2 bands, always use BIL scheme
if (len(dsDict.keys()) == 2
and os.path.splitext(infile)[1] not in ['.h5', '.he5']
and atr.get('scheme', 'BIL').upper() != 'BIL'):
print('the input binary file has 2 bands with band interleave as: {}'.
format(atr['scheme']))
print(
'for the output binary file, change the band interleave to BIL as default.'
)
atr['scheme'] = 'BIL'
if ext not in ['.h5', '.he5']:
writefile.write(dsDict,
out_file=outfile,
metadata=atr,
ref_file=infile)
return outfile
def calc_delay_timeseries(inps):
"""Calculate delay time-series and write it to HDF5 file.
Parameters: inps : namespace, all input parameters
Returns: tropo_file : str, file name of ECMWF.h5
"""
def get_dataset_size(fname):
atr = readfile.read_attribute(fname)
shape = (int(atr['LENGTH']), int(atr['WIDTH']))
return shape
def run_or_skip(grib_files, tropo_file, geom_file):
print('update mode: ON')
print('output file: {}'.format(tropo_file))
flag = 'skip'
# check existance and modification time
if ut.run_or_skip(out_file=tropo_file,
in_file=grib_files,
print_msg=False) == 'run':
flag = 'run'
print(
'1) output file either do NOT exist or is NOT newer than all GRIB files.'
)
else:
print('1) output file exists and is newer than all GRIB files.')
# check dataset size in space / time
date_list = [
str(re.findall('\d{8}', os.path.basename(i))[0])
for i in grib_files
]
if (get_dataset_size(tropo_file) != get_dataset_size(geom_file)
or any(i not in timeseries(tropo_file).get_date_list()
for i in date_list)):
flag = 'run'
print(
'2) output file does NOT have the same len/wid as the geometry file {} or does NOT contain all dates'
.format(geom_file))
else:
print(
'2) output file has the same len/wid as the geometry file and contains all dates'
)
# check if output file is fully written
with h5py.File(tropo_file, 'r') as f:
if np.all(f['timeseries'][-1, :, :] == 0):
flag = 'run'
print('3) output file is NOT fully written.')
else:
print('3) output file is fully written.')
# result
print('run or skip: {}'.format(flag))
return flag
if run_or_skip(inps.grib_files, inps.tropo_file, inps.geom_file) == 'skip':
return
## 1. prepare geometry data
geom_obj = geometry(inps.geom_file)
geom_obj.open()
inps.inc = geom_obj.read(datasetName='incidenceAngle')
inps.dem = geom_obj.read(datasetName='height')
# for testing
if inps.custom_height:
print(
'use input custom height of {} m for vertical integration'.format(
inps.custom_height))
inps.dem[:] = inps.custom_height
if 'latitude' in geom_obj.datasetNames:
# for lookup table in radar-coord (isce, doris)
inps.lat = geom_obj.read(datasetName='latitude')
inps.lon = geom_obj.read(datasetName='longitude')
elif 'Y_FIRST' in geom_obj.metadata:
# for lookup table in geo-coded (gamma, roipac) and obs. in geo-coord
inps.lat, inps.lon = ut.get_lat_lon(geom_obj.metadata)
# convert coordinates to lat/lon, e.g. from UTM for ASF HyPP3
if not geom_obj.metadata['Y_UNIT'].startswith('deg'):
inps.lat, inps.lon = ut.to_latlon(inps.atr['OG_FILE_PATH'],
inps.lon, inps.lat)
else:
# for lookup table in geo-coded (gamma, roipac) and obs. in radar-coord
inps.lat, inps.lon = ut.get_lat_lon_rdc(inps.atr)
# mask of valid pixels
mask = np.multiply(inps.inc != 0, ~np.isnan(inps.inc))
## 2. prepare output file
# metadata
atr = inps.atr.copy()
atr['FILE_TYPE'] = 'timeseries'
atr['UNIT'] = 'm'
# remove metadata related with double reference
# because absolute delay is calculated and saved
for key in ['REF_DATE', 'REF_X', 'REF_Y', 'REF_LAT', 'REF_LON']:
if key in atr.keys():
atr.pop(key)
# instantiate time-series
length, width = int(atr['LENGTH']), int(atr['WIDTH'])
num_date = len(inps.grib_files)
date_list = [
str(re.findall('\d{8}', os.path.basename(i))[0])
for i in inps.grib_files
]
dates = np.array(date_list, dtype=np.string_)
ds_name_dict = {
"date": [dates.dtype, (num_date, ), dates],
"timeseries": [np.float32, (num_date, length, width), None],
}
writefile.layout_hdf5(inps.tropo_file, ds_name_dict, metadata=atr)
## 3. calculate phase delay
print(
'\n------------------------------------------------------------------------------'
)
print(
'calculating absolute delay for each date using PyAPS (Jolivet et al., 2011; 2014) ...'
)
print('number of grib files used: {}'.format(num_date))
prog_bar = ptime.progressBar(maxValue=num_date, print_msg=~inps.verbose)
for i in range(num_date):
grib_file = inps.grib_files[i]
# calc tropo delay
tropo_data = get_delay(grib_file,
tropo_model=inps.tropo_model,
delay_type=inps.delay_type,
dem=inps.dem,
inc=inps.inc,
lat=inps.lat,
lon=inps.lon,
mask=mask,
verbose=inps.verbose)
# write tropo delay to file
block = [i, i + 1, 0, length, 0, width]
writefile.write_hdf5_block(inps.tropo_file,
data=tropo_data,
datasetName='timeseries',
block=block,
print_msg=False)
prog_bar.update(i + 1, suffix=os.path.basename(grib_file))
prog_bar.close()
return inps.tropo_file
def run_deramp(fname,
ramp_type,
mask_file=None,
out_file=None,
datasetName=None):
""" Remove ramp from each 2D matrix of input file
Parameters: fname : str, data file to be derampped
ramp_type : str, name of ramp to be estimated.
mask_file : str, file of mask of pixels used for ramp estimation
out_file : str, output file name
datasetName : str, output dataset name, for ifgramStack file type only
Returns: out_file : str, output file name
"""
start_time = time.time()
atr = readfile.read_attribute(fname)
k = atr['FILE_TYPE']
length = int(atr['LENGTH'])
width = int(atr['WIDTH'])
print('remove {} ramp from file: {}'.format(ramp_type, fname))
if not out_file:
fbase, fext = os.path.splitext(fname)
out_file = '{}_ramp{}'.format(fbase, fext)
if k == 'ifgramStack':
out_file = fname
# mask
if os.path.isfile(mask_file):
mask = readfile.read(mask_file)[0]
print('read mask file: ' + mask_file)
else:
mask = np.ones((length, width), dtype=np.bool_)
print('use mask of the whole area')
# deramping
if k == 'timeseries':
# write HDF5 file with defined metadata and (empty) dataset structure
writefile.layout_hdf5(out_file, ref_file=fname, print_msg=True)
print('estimating phase ramp one date at a time ...')
date_list = timeseries(fname).get_date_list()
num_date = len(date_list)
prog_bar = ptime.progressBar(maxValue=num_date)
for i in range(num_date):
# read
data = readfile.read(fname, datasetName=date_list[i])[0]
# deramp
data = deramp(data, mask, ramp_type=ramp_type, metadata=atr)[0]
# write
writefile.write_hdf5_block(out_file,
data,
datasetName='timeseries',
block=[i, i + 1, 0, length, 0, width],
print_msg=False)
prog_bar.update(i + 1, suffix='{}/{}'.format(i + 1, num_date))
prog_bar.close()
print('finished writing to file: {}'.format(out_file))
elif k == 'ifgramStack':
obj = ifgramStack(fname)
obj.open(print_msg=False)
if not datasetName:
datasetName = 'unwrapPhase'
with h5py.File(fname, 'a') as f:
ds = f[datasetName]
dsNameOut = '{}_ramp'.format(datasetName)
if dsNameOut in f.keys():
dsOut = f[dsNameOut]
print('access HDF5 dataset /{}'.format(dsNameOut))
else:
dsOut = f.create_dataset(dsNameOut,
shape=(obj.numIfgram, length, width),
dtype=np.float32,
chunks=True,
compression=None)
print('create HDF5 dataset /{}'.format(dsNameOut))
prog_bar = ptime.progressBar(maxValue=obj.numIfgram)
for i in range(obj.numIfgram):
data = ds[i, :, :]
data = deramp(data, mask, ramp_type=ramp_type, metadata=atr)[0]
dsOut[i, :, :] = data
prog_bar.update(i + 1,
suffix='{}/{}'.format(i + 1, obj.numIfgram))
prog_bar.close()
print('finished writing to file: {}'.format(fname))
# Single Dataset File
else:
data = readfile.read(fname)[0]
data = deramp(data, mask, ramp_type, metadata=atr)[0]
print('writing >>> {}'.format(out_file))
writefile.write(data, out_file=out_file, ref_file=fname)
m, s = divmod(time.time() - start_time, 60)
print('time used: {:02.0f} mins {:02.1f} secs.'.format(m, s))
return out_file
def prepare_stack(outfile,
unw_file,
metadata,
processor,
baseline_dir=None,
box=None):
print('-' * 50)
print('preparing ifgramStack file: {}'.format(outfile))
# copy metadata to meta
meta = {key: value for key, value in metadata.items()}
# get list of *.unw file
unw_files = sorted(glob.glob(unw_file))
num_pair = len(unw_files)
print('number of interferograms:', num_pair)
# get list of *.unw.conncomp file
cc_files = [f'{x}.conncomp' for x in unw_files]
cc_files = [x for x in cc_files if os.path.isfile(x)]
print(f'number of connected components files: {len(cc_files)}')
if len(cc_files) != len(unw_files):
print(
'the number of *.unw and *.unw.conncomp files are NOT consistent')
print('skip creating ifgramStack.h5 file.')
return
# get date info: date12_list
date12_list = [os.path.basename(x).split('.')[0] for x in unw_files]
# prepare baseline info
if baseline_dir is not None:
# read baseline timeseries
baseline_dict = isce_utils.read_baseline_timeseries(
baseline_dir, processor=processor)
# calc baseline for each pair
print('calc perp baseline pairs from time-series')
pbase = np.zeros(num_pair, dtype=np.float32)
for i, date12 in enumerate(date12_list):
[date1, date2] = date12.split('_')
pbase[i] = np.subtract(baseline_dict[date2],
baseline_dict[date1]).mean()
# size info
box = box if box else (0, 0, int(meta['WIDTH']), int(meta['LENGTH']))
kwargs = dict(xoff=box[0],
yoff=box[1],
win_xsize=box[2] - box[0],
win_ysize=box[3] - box[1])
# define (and fill out some) dataset structure
date12_arr = np.array([x.split('_') for x in date12_list],
dtype=np.string_)
drop_ifgram = np.ones(num_pair, dtype=np.bool_)
ds_name_dict = {
"date": [date12_arr.dtype, (num_pair, 2), date12_arr],
"bperp": [np.float32, (num_pair, ), pbase],
"dropIfgram": [np.bool_, (num_pair, ), drop_ifgram],
"unwrapPhase":
[np.float32, (num_pair, box[3] - box[1], box[2] - box[0]), None],
"connectComponent":
[np.float32, (num_pair, box[3] - box[1], box[2] - box[0]), None],
}
# initiate HDF5 file
meta["FILE_TYPE"] = "ifgramStack"
writefile.layout_hdf5(outfile, ds_name_dict, metadata=meta)
# writing data to HDF5 file
print('writing data to HDF5 file {} with a mode ...'.format(outfile))
with h5py.File(outfile, "a") as f:
prog_bar = ptime.progressBar(maxValue=num_pair)
for i, (unw_file, cc_file) in enumerate(zip(unw_files, cc_files)):
# read/write *.unw file
ds = gdal.Open(unw_file, gdal.GA_ReadOnly)
data = np.array(ds.GetRasterBand(2).ReadAsArray(**kwargs),
dtype=np.float32)
f["unwrapPhase"][i] = data
# read/write *.unw.conncomp file
ds = gdal.Open(cc_file, gdal.GA_ReadOnly)
data = np.array(ds.GetRasterBand(1).ReadAsArray(**kwargs),
dtype=np.float32)
f["connectComponent"][i] = data
prog_bar.update(i + 1, suffix=date12_list[i])
prog_bar.close()
print('finished writing to HDF5 file: {}'.format(outfile))
return outfile
def diff_file(file1, file2, out_file=None, force=False, max_num_pixel=2e8):
"""calculate/write file1 - file2
Parameters: file1 - str, path of file1
file2 - list of str, path of file2(s)
out_file - str, path of output file
force - bool, overwrite existing output file
max_num_pixel - float, maximum number of pixels for each block
"""
start_time = time.time()
if not out_file:
fbase, fext = os.path.splitext(file1)
if len(file2) > 1:
raise ValueError(
'Output file name is needed for more than 2 files input.')
out_file = '{}_diff_{}{}'.format(
fbase,
os.path.splitext(os.path.basename(file2[0]))[0], fext)
print('{} - {} --> {}'.format(file1, file2, out_file))
# Read basic info
atr1 = readfile.read_attribute(file1)
k1 = atr1['FILE_TYPE']
atr2 = readfile.read_attribute(file2[0])
k2 = atr2['FILE_TYPE']
print('input files are: {} and {}'.format(k1, k2))
if k1 == 'timeseries':
if k2 not in ['timeseries', 'giantTimeseries']:
raise Exception(
'Input multiple dataset files are not the same file type!')
if len(file2) > 1:
raise Exception(
('Only 2 files substraction is supported for time-series file,'
' {} input.'.format(len(file2) + 1)))
atr1 = readfile.read_attribute(file1)
atr2 = readfile.read_attribute(file2[0])
dateList1 = timeseries(file1).get_date_list()
if k2 == 'timeseries':
dateList2 = timeseries(file2[0]).get_date_list()
unit_fac = 1.
elif k2 == 'giantTimeseries':
dateList2 = giantTimeseries(file2[0]).get_date_list()
unit_fac = 0.001
# check reference point
ref_date, ref_y, ref_x = check_reference(atr1, atr2)
# check dates shared by two timeseries files
dateListShared = [i for i in dateList1 if i in dateList2]
dateShared = np.ones((len(dateList1)), dtype=np.bool_)
if dateListShared != dateList1:
print('WARNING: {} does not contain all dates in {}'.format(
file2, file1))
if force:
dateListEx = list(set(dateList1) - set(dateListShared))
print(
'Continue and enforce the differencing for their shared dates only.'
)
print(
'\twith following dates are ignored for differencing:\n{}'.
format(dateListEx))
dateShared[np.array([dateList1.index(i)
for i in dateListEx])] = 0
else:
raise Exception(
'To enforce the differencing anyway, use --force option.')
# instantiate the output file
writefile.layout_hdf5(out_file, ref_file=file1)
# block-by-block IO
length, width = int(atr1['LENGTH']), int(atr1['WIDTH'])
num_box = int(np.ceil(len(dateList1) * length * width / max_num_pixel))
box_list = cluster.split_box2sub_boxes(box=(0, 0, width, length),
num_split=num_box,
dimension='y',
print_msg=True)
if ref_y and ref_x:
ref_box = (ref_x, ref_y, ref_x + 1, ref_y + 1)
ref_val = readfile.read(file2[0],
datasetName=dateListShared,
box=ref_box)[0] * unit_fac
for i, box in enumerate(box_list):
if num_box > 1:
print('\n------- processing patch {} out of {} --------------'.
format(i + 1, num_box))
print('box: {}'.format(box))
# read data2 (consider different reference_date/pixel)
print('read from file: {}'.format(file2[0]))
data2 = readfile.read(
file2[0], datasetName=dateListShared, box=box)[0] * unit_fac
if ref_y and ref_x:
print('* referencing data from {} to y/x: {}/{}'.format(
os.path.basename(file2[0]), ref_y, ref_x))
data2 -= np.tile(ref_val.reshape(-1, 1, 1),
(1, data2.shape[1], data2.shape[2]))
if ref_date:
print('* referencing data from {} to date: {}'.format(
os.path.basename(file2[0]), ref_date))
ref_ind = dateListShared.index(ref_date)
data2 -= np.tile(data2[ref_ind, :, :], (data2.shape[0], 1, 1))
# read data1
print('read from file: {}'.format(file1))
data = readfile.read(file1, box=box)[0]
# apply differencing
mask = data == 0.
data[dateShared] -= data2
data[mask] = 0. # Do not change zero phase value
del data2
# write the block
block = [0, data.shape[0], box[1], box[3], box[0], box[2]]
writefile.write_hdf5_block(out_file,
data=data,
datasetName=k1,
block=block)
elif all(i == 'ifgramStack' for i in [k1, k2]):
obj1 = ifgramStack(file1)
obj1.open()
obj2 = ifgramStack(file2[0])
obj2.open()
dsNames = list(set(obj1.datasetNames) & set(obj2.datasetNames))
if len(dsNames) == 0:
raise ValueError('no common dataset between two files!')
dsName = [i for i in ifgramDatasetNames if i in dsNames][0]
# read data
print('reading {} from file {} ...'.format(dsName, file1))
data1 = readfile.read(file1, datasetName=dsName)[0]
print('reading {} from file {} ...'.format(dsName, file2[0]))
data2 = readfile.read(file2[0], datasetName=dsName)[0]
# consider reference pixel
if 'unwrapphase' in dsName.lower():
print('referencing to pixel ({},{}) ...'.format(
obj1.refY, obj1.refX))
ref1 = data1[:, obj1.refY, obj1.refX]
ref2 = data2[:, obj2.refY, obj2.refX]
for i in range(data1.shape[0]):
data1[i, :][data1[i, :] != 0.] -= ref1[i]
data2[i, :][data2[i, :] != 0.] -= ref2[i]
# operation and ignore zero values
data1[data1 == 0] = np.nan
data2[data2 == 0] = np.nan
data = data1 - data2
del data1, data2
data[np.isnan(data)] = 0.
# write to file
dsDict = {}
dsDict[dsName] = data
writefile.write(dsDict, out_file=out_file, ref_file=file1)
# Sing dataset file
else:
data1 = readfile.read(file1)[0]
data = np.array(data1, data1.dtype)
for fname in file2:
data2 = readfile.read(fname)[0]
data = np.array(data, dtype=np.float32) - np.array(
data2, dtype=np.float32)
data = np.array(data, data1.dtype)
print('writing >>> ' + out_file)
writefile.write(data, out_file=out_file, metadata=atr1)
m, s = divmod(time.time() - start_time, 60)
print('time used: {:02.0f} mins {:02.1f} secs'.format(m, s))
return out_file
def main(iargs=None):
inps = cmd_line_parse(iargs)
stack_obj = ifgramStack(inps.ifgram_stack)
stack_obj.open()
length, width = stack_obj.length, stack_obj.width
date12_list = stack_obj.get_date12_list(dropIfgram=True)
date12_list_all = stack_obj.get_date12_list(dropIfgram=False)
print('scene length, width', length, width)
ref_phase = stack_obj.get_reference_phase(unwDatasetName='unwrapPhase')
inps.length = length
inps.width = width
# retrieve the list of SLC dates from ifgramStack.h5
ifgram0 = date12_list[0]
date1, date2 = ifgram0.split('_')
SLC_list = [date1, date2]
for ifgram in date12_list:
date1, date2 = ifgram.split('_')
if date1 not in SLC_list:
SLC_list.append(date1)
if date2 not in SLC_list:
SLC_list.append(date2)
SLC_list.sort()
print('number of SLC found : ', len(SLC_list))
print('first SLC: ', SLC_list[0])
print('last SLC: ', SLC_list[-1])
# split igram_file into blocks to save memory
box_list, num_box = ifginv.split2boxes(inps.ifgram_stack, inps.max_memory)
closurephase = np.zeros([length, width], np.complex64)
#process block-by-block
for i, box in enumerate(box_list):
box_width = box[2] - box[0]
box_length = box[3] - box[1]
print(box)
if num_box > 1:
print('\n------- processing patch {} out of {} --------------'.
format(i + 1, num_box))
print('box width: {}'.format(box_width))
print('box length: {}'.format(box_length))
closurephase[box[1]:box[3],
box[0]:box[2]], numcp = cum_seq_closurePhase(
SLC_list, date12_list_all, inps.ifgram_stack,
ref_phase, inps.nl, box)
# What is a good thredshold?
# Assume that it's pure noise so that the phase is uniform distributed from -pi to pi.
# The standard deviation of phase in each loop is pi/sqrt(3) (technically should be smaller because when forming loops there should be a reduction in phase variance)
# The standard deviation of phase in cumulative wrapped closure phase is pi/sqrt(3)/sqrt(num_cp) -- again another simplification assuming no correlation.
# We use 3\delta as default threshold -- 99.7% confidence
if inps.numsigma:
threshold_cp = np.pi / np.sqrt(3) / np.sqrt(numcp) * inps.numsigma
else:
threshold_cp = np.pi / np.sqrt(3) / np.sqrt(
numcp) * 3 # 3/sigma, 99.7% confidence
mask = np.ones([length, width], np.float32)
mask[np.abs(np.angle(closurephase)) >
threshold_cp] = 0 # this masks areas with potential bias
mask[np.abs(
np.abs(closurephase) / numcp < inps.episilon
)] = 1 # this unmasks areas with low correlation (where it's hard to know wheter there is bias either)
# save mask
meta = dict(stack_obj.metadata)
meta['FILE_TYPE'] = 'mask'
ds_name_dict = {
'cpmask': [np.float32, (length, width), mask],
}
writefile.layout_hdf5(os.path.join(inps.outdir, 'cpmask.h5'), ds_name_dict,
meta)
# also save the average closure phase
ds_name_dict2 = {
'phase': [np.float32, (length, width),
np.angle(closurephase)],
'amplitude':
[np.float32, (length, width),
np.abs(closurephase) / numcp],
}
writefile.layout_hdf5(os.path.join(inps.outdir, 'avgwcp.h5'),
ds_name_dict2, meta)
def correct_dem_error(inps):
"""Correct DEM error of input timeseries file"""
start_time = time.time()
## 1. input info
# 1.1 read date info
ts_obj = timeseries(inps.timeseries_file)
ts_obj.open()
num_date = ts_obj.numDate
length, width = ts_obj.length, ts_obj.width
num_step = len(inps.stepFuncDate)
# exclude dates
date_flag = read_exclude_date(inps.excludeDate, ts_obj.dateList)[0]
if inps.polyOrder > np.sum(date_flag):
raise ValueError(
"input poly order {} > number of acquisition {}! Reduce it!".
format(inps.polyOrder, np.sum(date_flag)))
# 1.2 design matrix part 1 - time func for surface deformation
G_defo = get_design_matrix4defo(inps)
## 2. prepare output
# 2.1 metadata
meta = dict(ts_obj.metadata)
print(
'add/update the following configuration metadata to file:\n{}'.format(
configKeys))
for key in configKeys:
meta[key_prefix + key] = str(vars(inps)[key])
# 2.2 instantiate est. DEM error
dem_err_file = 'demErr.h5'
meta['FILE_TYPE'] = 'dem'
meta['UNIT'] = 'm'
ds_name_dict = {'dem': [np.float32, (length, width), None]}
writefile.layout_hdf5(dem_err_file, ds_name_dict, metadata=meta)
# 2.3 instantiate corrected time-series
ts_cor_file = inps.outfile
meta['FILE_TYPE'] = 'timeseries'
writefile.layout_hdf5(ts_cor_file,
metadata=meta,
ref_file=inps.timeseries_file)
# 2.4 instantiate residual phase time-series
ts_res_file = os.path.join(os.path.dirname(inps.outfile),
'timeseriesResidual.h5')
writefile.layout_hdf5(ts_res_file,
metadata=meta,
ref_file=inps.timeseries_file)
# 2.5 instantiate est. step model(s)
step_file = None
if num_step > 0:
step_file = os.path.join(os.path.dirname(inps.outfile),
'timeseriesStepModel.h5')
meta.pop('REF_DATE')
step_dates = np.array(inps.stepFuncDate, dtype=np.string_)
ds_name_dict = {
'date': [step_dates.dtype, (num_step, ), step_dates],
'timeseries': [np.float32, (num_step, length, width), None]
}
writefile.layout_hdf5(step_file, ds_name_dict, metadata=meta)
## 3. run the estimation and write to disk
# 3.1 split ts_file into blocks to save memory
box_list, num_box = split2boxes(inps.timeseries_file,
geom_file=inps.geom_file,
memory_size=inps.memorySize,
num_step=num_step)
# 3.2 invert / write block-by-block
for i, box in enumerate(box_list):
box_width = box[2] - box[0]
box_length = box[3] - box[1]
if num_box > 1:
print('\n------- processing patch {} out of {} --------------'.
format(i + 1, num_box))
print('box width: {}'.format(box_width))
print('box length: {}'.format(box_length))
# invert
(delta_z, ts_cor, ts_res, step_model) = correct_dem_error_patch(
G_defo,
ts_file=inps.timeseries_file,
geom_file=inps.geom_file,
box=box,
date_flag=date_flag,
num_step=num_step,
phase_velocity=inps.phaseVelocity)
# write the block to disk
# with 3D block in [z0, z1, y0, y1, x0, x1]
# and 2D block in [y0, y1, x0, x1]
# DEM error - 2D
block = [box[1], box[3], box[0], box[2]]
writefile.write_hdf5_block(dem_err_file,
data=delta_z,
datasetName='dem',
block=block)
# corrected time-series - 3D
block = [0, num_date, box[1], box[3], box[0], box[2]]
writefile.write_hdf5_block(ts_cor_file,
data=ts_cor,
datasetName='timeseries',
block=block)
# residual time-series - 3D
block = [0, num_date, box[1], box[3], box[0], box[2]]
writefile.write_hdf5_block(ts_res_file,
data=ts_res,
datasetName='timeseries',
block=block)
# step func time-series - 3D
if num_step > 0:
block = [0, num_step, box[1], box[3], box[0], box[2]]
writefile.write_hdf5_block(step_file,
data=step_model,
datasetName='timeseries',
block=block)
# time info
m, s = divmod(time.time() - start_time, 60)
print('time used: {:02.0f} mins {:02.1f} secs.'.format(m, s))
return dem_err_file, ts_cor_file, ts_res_file, step_file
def run_timeseries2time_func(inps):
# basic info
atr = readfile.read_attribute(inps.timeseries_file)
length, width = int(atr['LENGTH']), int(atr['WIDTH'])
num_date = inps.numDate
dates = np.array(inps.dateList)
seconds = atr.get('CENTER_LINE_UTC', 0)
# use the 1st date as reference if not found, e.g. timeseriesResidual.h5 file
if "REF_DATE" not in atr.keys() and not inps.ref_date:
inps.ref_date = inps.dateList[0]
print(
'WARNING: No REF_DATE found in time-series file or input in command line.'
)
print(' Set "--ref-date {}" and continue.'.format(inps.dateList[0]))
# get deformation model from parsers
model, num_param = read_inps2model(inps)
## output preparation
# time_func_param: attributes
atrV = dict(atr)
atrV['FILE_TYPE'] = 'velocity'
atrV['UNIT'] = 'm/year'
atrV['START_DATE'] = inps.dateList[0]
atrV['END_DATE'] = inps.dateList[-1]
atrV['DATE12'] = '{}_{}'.format(inps.dateList[0], inps.dateList[-1])
if inps.ref_yx:
atrV['REF_Y'] = inps.ref_yx[0]
atrV['REF_X'] = inps.ref_yx[1]
if inps.ref_date:
atrV['REF_DATE'] = inps.ref_date
# time_func_param: config parameter
print('add/update the following configuration metadata:\n{}'.format(
configKeys))
for key in configKeys:
atrV[key_prefix + key] = str(vars(inps)[key])
# time_func_param: instantiate output file
ds_name_dict, ds_unit_dict = model2hdf5_dataset(model,
ds_shape=(length,
width))[1:]
writefile.layout_hdf5(inps.outfile,
metadata=atrV,
ds_name_dict=ds_name_dict,
ds_unit_dict=ds_unit_dict)
# timeseries_res: attributes + instantiate output file
if inps.save_res:
atrR = dict(atr)
# remove REF_DATE attribute
for key in ['REF_DATE']:
if key in atrR.keys():
atrR.pop(key)
# prepare ds_name_dict manually, instead of using ref_file, to support --ex option
date_len = len(inps.dateList[0])
ds_name_dict = {
"date": [
np.dtype(f'S{date_len}'), (num_date, ),
np.array(inps.dateList, dtype=np.string_)
],
"timeseries": [np.float32, (num_date, length, width), None]
}
writefile.layout_hdf5(inps.res_file,
ds_name_dict=ds_name_dict,
metadata=atrR)
## estimation
# calc number of box based on memory limit
memoryAll = (num_date + num_param * 2 + 2) * length * width * 4
if inps.bootstrap:
memoryAll += inps.bootstrapCount * num_param * length * width * 4
num_box = int(np.ceil(memoryAll * 3 / (inps.maxMemory * 1024**3)))
box_list = cluster.split_box2sub_boxes(box=(0, 0, width, length),
num_split=num_box,
dimension='y',
print_msg=True)
# loop for block-by-block IO
for i, box in enumerate(box_list):
box_wid = box[2] - box[0]
box_len = box[3] - box[1]
num_pixel = box_len * box_wid
if num_box > 1:
print('\n------- processing patch {} out of {} --------------'.
format(i + 1, num_box))
print('box width: {}'.format(box_wid))
print('box length: {}'.format(box_len))
# initiate output
m = np.zeros((num_param, num_pixel), dtype=dataType)
m_std = np.zeros((num_param, num_pixel), dtype=dataType)
# read input
print('reading data from file {} ...'.format(inps.timeseries_file))
ts_data = readfile.read(inps.timeseries_file, box=box)[0]
# referencing in time and space
# for file w/o reference info. e.g. ERA5.h5
if inps.ref_date:
print('referecing to date: {}'.format(inps.ref_date))
ref_ind = inps.dateList.index(inps.ref_date)
ts_data -= np.tile(ts_data[ref_ind, :, :],
(ts_data.shape[0], 1, 1))
if inps.ref_yx:
print('referencing to point (y, x): ({}, {})'.format(
inps.ref_yx[0], inps.ref_yx[1]))
ref_box = (inps.ref_yx[1], inps.ref_yx[0], inps.ref_yx[1] + 1,
inps.ref_yx[0] + 1)
ref_val = readfile.read(inps.timeseries_file, box=ref_box)[0]
ts_data -= np.tile(ref_val.reshape(ts_data.shape[0], 1, 1),
(1, ts_data.shape[1], ts_data.shape[2]))
ts_data = ts_data[inps.dropDate, :, :].reshape(inps.numDate, -1)
if atrV['UNIT'] == 'mm':
ts_data *= 1. / 1000.
ts_cov = None
if inps.ts_cov_file:
print(
f'reading time-series covariance matrix from file {inps.ts_cov_file} ...'
)
ts_cov = readfile.read(inps.ts_cov_file, box=box)[0]
if len(ts_cov.shape) == 4:
# full covariance matrix in 4D --> 3D
if inps.numDate < ts_cov.shape[0]:
ts_cov = ts_cov[inps.dropDate, :, :, :]
ts_cov = ts_cov[:, inps.dropDate, :, :]
ts_cov = ts_cov.reshape(inps.numDate, inps.numDate, -1)
elif len(ts_cov.shape) == 3:
# diaginal variance matrix in 3D --> 2D
if inps.numDate < ts_cov.shape[0]:
ts_cov = ts_cov[inps.dropDate, :, :]
ts_cov = ts_cov.reshape(inps.numDate, -1)
## set zero value to a fixed small value to avoid divide by zero
#epsilon = 1e-5
#ts_cov[ts_cov<epsilon] = epsilon
# mask invalid pixels
print('skip pixels with zero/nan value in all acquisitions')
ts_stack = np.nanmean(ts_data, axis=0)
mask = np.multiply(~np.isnan(ts_stack), ts_stack != 0.)
del ts_stack
#if ts_cov is not None:
# print('skip pxiels with nan STD value in any acquisition')
# num_std_nan = np.sum(np.isnan(ts_cov), axis=0)
# mask *= num_std_nan == 0
# del num_std_nan
ts_data = ts_data[:, mask]
num_pixel2inv = int(np.sum(mask))
idx_pixel2inv = np.where(mask)[0]
print('number of pixels to invert: {} out of {} ({:.1f}%)'.format(
num_pixel2inv, num_pixel, num_pixel2inv / num_pixel * 100))
# go to next if no valid pixel found
if num_pixel2inv == 0:
continue
### estimation / solve Gm = d
print('estimating time functions via linalg.lstsq ...')
if inps.bootstrap:
## option 1 - least squares with bootstrapping
# Bootstrapping is a resampling method which can be used to estimate properties
# of an estimator. The method relies on independently sampling the data set with
# replacement.
print(
'estimating time function STD with bootstrap resampling ({} times) ...'
.format(inps.bootstrapCount))
# calc model of all bootstrap sampling
rng = np.random.default_rng()
m_boot = np.zeros((inps.bootstrapCount, num_param, num_pixel2inv),
dtype=dataType)
prog_bar = ptime.progressBar(maxValue=inps.bootstrapCount)
for i in range(inps.bootstrapCount):
# bootstrap resampling
boot_ind = rng.choice(inps.numDate,
size=inps.numDate,
replace=True)
boot_ind.sort()
# estimation
m_boot[i] = time_func.estimate_time_func(
model=model,
date_list=dates[boot_ind].tolist(),
dis_ts=ts_data[boot_ind],
seconds=seconds)[1]
prog_bar.update(i + 1,
suffix='iteration {} / {}'.format(
i + 1, inps.bootstrapCount))
prog_bar.close()
#del ts_data
# get mean/std among all bootstrap sampling
m[:, mask] = m_boot.mean(axis=0).reshape(num_param, -1)
m_std[:, mask] = m_boot.std(axis=0).reshape(num_param, -1)
del m_boot
# get design matrix to calculate the residual time series
G = time_func.get_design_matrix4time_func(inps.dateList,
model=model,
ref_date=inps.ref_date,
seconds=seconds)
else:
## option 2 - least squares with uncertainty propagation
G, m[:, mask], e2 = time_func.estimate_time_func(
model=model,
date_list=inps.dateList,
dis_ts=ts_data,
seconds=seconds)
#del ts_data
## Compute the covariance matrix for model parameters:
# G * m = d
# C_m_hat = G+ * C_d * G+.T
#
# For ordinary least squares estimation:
# G+ = (G.T * G)^-1 * G.T (option 2.1)
#
# For weighted least squares estimation:
# G+ = (G.T * C_d^-1 * G)^-1 * G.T * C_d^-1
# => C_m_hat = (G.T * C_d^-1 * G)^-1 (option 2.2)
#
# Assuming normality of the observation errors (in the time domain) with a variance of sigma^2
# we have C_d = sigma^2 * I, then the above equation is simplfied into:
# C_m_hat = sigma^2 * (G.T * G)^-1 (option 2.3)
#
# Based on the law of integrated expectation, we estimate the obs sigma^2 using
# the OLS estimation residual as:
# e_hat = d - d_hat
# => sigma_hat^2 = (e_hat.T * e_hat) / N
# => sigma^2 = sigma_hat^2 * N / (N - P) (option 2.4)
# = (e_hat.T * e_hat) / (N - P)
# which is the equation (10) from Fattahi and Amelung (2015, JGR)
if ts_cov is not None:
# option 2.1 - linear propagation from time-series (co)variance matrix
# TO DO: save the full covariance matrix of the time function parameters
# only the STD is saved right now
covar_flag = True if len(ts_cov.shape) == 3 else False
msg = 'estimating time function STD from time-serries '
msg += 'covariance pixel-by-pixel ...' if covar_flag else 'variance pixel-by-pixel ...'
print(msg)
# calc the common pseudo-inverse matrix
Gplus = linalg.pinv(G)
# loop over each pixel
# or use multidimension matrix multiplication
# m_cov = Gplus @ ts_cov @ Gplus.T
prog_bar = ptime.progressBar(maxValue=num_pixel2inv)
for i in range(num_pixel2inv):
idx = idx_pixel2inv[i]
# cov: time-series -> time func
ts_covi = ts_cov[:, :, idx] if covar_flag else np.diag(
ts_cov[:, idx])
m_cov = np.linalg.multi_dot([Gplus, ts_covi, Gplus.T])
m_std[:, idx] = np.sqrt(np.diag(m_cov))
prog_bar.update(i + 1,
every=200,
suffix='{}/{} pixels'.format(
i + 1, num_pixel2inv))
prog_bar.close()
else:
# option 2.3 - assume obs errors following normal dist. in time
print(
'estimating time function STD from time-series fitting residual ...'
)
G_inv = linalg.inv(np.dot(G.T, G))
m_var = e2.reshape(1, -1) / (num_date - num_param)
m_std[:, mask] = np.sqrt(
np.dot(np.diag(G_inv).reshape(-1, 1), m_var))
# option 2.4 - simplified form for linear velocity (without matrix linear algebra)
# The STD can also be calculated using Eq. (10) from Fattahi and Amelung (2015, JGR)
# ts_diff = ts_data - np.dot(G, m)
# t_diff = G[:, 1] - np.mean(G[:, 1])
# vel_std = np.sqrt(np.sum(ts_diff ** 2, axis=0) / np.sum(t_diff ** 2) / (num_date - 2))
# write - time func params
block = [box[1], box[3], box[0], box[2]]
ds_dict = model2hdf5_dataset(model, m, m_std, mask=mask)[0]
for ds_name, data in ds_dict.items():
writefile.write_hdf5_block(inps.outfile,
data=data.reshape(box_len, box_wid),
datasetName=ds_name,
block=block)
# write - residual file
if inps.save_res:
block = [0, num_date, box[1], box[3], box[0], box[2]]
ts_res = np.ones(
(num_date, box_len * box_wid), dtype=np.float32) * np.nan
ts_res[:, mask] = ts_data - np.dot(G, m)[:, mask]
writefile.write_hdf5_block(inps.res_file,
data=ts_res.reshape(
num_date, box_len, box_wid),
datasetName='timeseries',
block=block)
return inps.outfile
def layout_hdf5(out_file, atr, model):
"""create HDF5 file for estimated time functions
with defined metadata and (empty) dataset structure
"""
# deformation model info
poly_deg = model['polynomial']
num_period = len(model['periodic'])
num_step = len(model['step'])
# size info
length = int(atr['LENGTH'])
width = int(atr['WIDTH'])
ds_name_dict = {}
ds_unit_dict = {}
# time func 1 - polynomial
for i in range(1, poly_deg + 1):
# dataset name
if i == 1:
dsName = 'velocity'
unit = 'm/year'
elif i == 2:
dsName = 'acceleration'
unit = 'm/year^2'
else:
dsName = 'poly{}'.format(i)
unit = 'm/year^{}'.format(i)
# update ds_name/unit_dict
ds_name_dict[dsName] = [dataType, (length, width), None]
ds_unit_dict[dsName] = unit
ds_name_dict[dsName+'Std'] = [dataType, (length, width), None]
ds_unit_dict[dsName+'Std'] = unit
# time func 2 - periodic
for i in range(num_period):
# dataset name
period = model['periodic'][i]
if period == 1:
dsName = 'annualAmp'
elif period == 0.5:
dsName = 'semiAnnualAmp'
else:
dsName = 'periodY{}Amp'.format(period)
# update ds_name/unit_dict
ds_name_dict[dsName] = [dataType, (length, width), None]
ds_unit_dict[dsName] = 'm'
# time func 3 - step
for i in range(num_step):
# dataset name
dsName = 'step{}'.format(model['step'][i])
# update ds_name/unit_dict
ds_name_dict[dsName] = [dataType, (length, width), None]
ds_unit_dict[dsName] = 'm'
ds_name_dict[dsName+'Std'] = [dataType, (length, width), None]
ds_unit_dict[dsName+'Std'] = 'm'
# layout hdf5
writefile.layout_hdf5(out_file, ds_name_dict, metadata=atr)
# add metadata to HDF5 dataset
max_digit = max([len(i) for i in ds_unit_dict.keys()])
with h5py.File(out_file, 'r+') as f:
for key, value in ds_unit_dict.items():
f[key].attrs['UNIT'] = value
print('add /{d:<{w}} attribute: UNIT = {u}'.format(d=key,
w=max_digit,
u=value))
return out_file
def correct_dem_error(inps):
"""Correct DEM error of input timeseries file"""
start_time = time.time()
# limit the number of threads to 1
# for slight speedup and big CPU usage save
num_threads_dict = cluster.set_num_threads("1")
## 1. input info
# 1.1 read date info
ts_obj = timeseries(inps.timeseries_file)
ts_obj.open()
num_date = ts_obj.numDate
length, width = ts_obj.length, ts_obj.width
num_step = len(inps.stepFuncDate)
# exclude dates
date_flag = read_exclude_date(inps.excludeDate, ts_obj.dateList)[0]
if inps.polyOrder > np.sum(date_flag):
raise ValueError(
"input poly order {} > number of acquisition {}! Reduce it!".
format(inps.polyOrder, np.sum(date_flag)))
# 1.2 design matrix part 1 - time func for surface deformation
G_defo = get_design_matrix4defo(inps)
## 2. prepare output
# 2.1 metadata
meta = dict(ts_obj.metadata)
print(
'add/update the following configuration metadata to file:\n{}'.format(
configKeys))
for key in configKeys:
meta[key_prefix + key] = str(vars(inps)[key])
# 2.2 instantiate est. DEM error
dem_err_file = 'demErr.h5'
meta['FILE_TYPE'] = 'dem'
meta['UNIT'] = 'm'
ds_name_dict = {'dem': [np.float32, (length, width), None]}
writefile.layout_hdf5(dem_err_file, ds_name_dict, metadata=meta)
# 2.3 instantiate corrected time-series
ts_cor_file = inps.outfile
meta['FILE_TYPE'] = 'timeseries'
writefile.layout_hdf5(ts_cor_file,
metadata=meta,
ref_file=inps.timeseries_file)
# 2.4 instantiate residual phase time-series
ts_res_file = os.path.join(os.path.dirname(inps.outfile),
'timeseriesResidual.h5')
writefile.layout_hdf5(ts_res_file,
metadata=meta,
ref_file=inps.timeseries_file)
## 3. run the estimation and write to disk
# 3.1 split ts_file into blocks to save memory
# 1st dimension size: ts (obs / cor / res / step) + dem_err/inc_angle/rg_dist (+pbase)
num_epoch = num_date * 3 + num_step + 3
if inps.geom_file:
geom_obj = geometry(inps.geom_file)
geom_obj.open(print_msg=False)
if 'bperp' in geom_obj.datasetNames:
num_epoch += num_date
# split in row/line direction based on the input memory limit
num_box = int(
np.ceil((num_epoch * length * width * 4) * 2.5 /
(inps.maxMemory * 1024**3)))
box_list = cluster.split_box2sub_boxes(box=(0, 0, width, length),
num_split=num_box,
dimension='y')
# 3.2 prepare the input arguments for *_patch()
data_kwargs = {
'G_defo': G_defo,
'ts_file': inps.timeseries_file,
'geom_file': inps.geom_file,
'date_flag': date_flag,
'phase_velocity': inps.phaseVelocity,
}
# 3.3 invert / write block-by-block
for i, box in enumerate(box_list):
box_wid = box[2] - box[0]
box_len = box[3] - box[1]
if num_box > 1:
print('\n------- processing patch {} out of {} --------------'.
format(i + 1, num_box))
print('box width: {}'.format(box_wid))
print('box length: {}'.format(box_len))
# update box argument in the input data
data_kwargs['box'] = box
# invert
if not inps.cluster:
# non-parallel
delta_z, ts_cor, ts_res = correct_dem_error_patch(
**data_kwargs)[:-1]
else:
# parallel
print('\n\n------- start parallel processing using Dask -------')
# initiate the output data
delta_z = np.zeros((box_len, box_wid), dtype=np.float32)
ts_cor = np.zeros((num_date, box_len, box_wid), dtype=np.float32)
ts_res = np.zeros((num_date, box_len, box_wid), dtype=np.float32)
# initiate dask cluster and client
cluster_obj = cluster.DaskCluster(inps.cluster,
inps.numWorker,
config_name=inps.config)
cluster_obj.open()
# run dask
delta_z, ts_cor, ts_res = cluster_obj.run(
func=correct_dem_error_patch,
func_data=data_kwargs,
results=[delta_z, ts_cor, ts_res])
# close dask cluster and client
cluster_obj.close()
print('------- finished parallel processing -------\n\n')
# write the block to disk
# with 3D block in [z0, z1, y0, y1, x0, x1]
# and 2D block in [y0, y1, x0, x1]
# DEM error - 2D
block = [box[1], box[3], box[0], box[2]]
writefile.write_hdf5_block(dem_err_file,
data=delta_z,
datasetName='dem',
block=block)
# corrected time-series - 3D
block = [0, num_date, box[1], box[3], box[0], box[2]]
writefile.write_hdf5_block(ts_cor_file,
data=ts_cor,
datasetName='timeseries',
block=block)
# residual time-series - 3D
block = [0, num_date, box[1], box[3], box[0], box[2]]
writefile.write_hdf5_block(ts_res_file,
data=ts_res,
datasetName='timeseries',
block=block)
# roll back to the origial number of threads
cluster.roll_back_num_threads(num_threads_dict)
# time info
m, s = divmod(time.time() - start_time, 60)
print('time used: {:02.0f} mins {:02.1f} secs.'.format(m, s))
return dem_err_file, ts_cor_file, ts_res_file
def run_timeseries2time_func(inps):
# basic info
atr = readfile.read_attribute(inps.timeseries_file)
length, width = int(atr['LENGTH']), int(atr['WIDTH'])
num_date = inps.numDate
dates = np.array(inps.dateList)
seconds = atr.get('CENTER_LINE_UTC', 0)
# use the 1st date as reference if not found, e.g. timeseriesResidual.h5 file
if "REF_DATE" not in atr.keys() and not inps.ref_date:
inps.ref_date = inps.dateList[0]
print(
'WARNING: No REF_DATE found in time-series file or input in command line.'
)
print(' Set "--ref-date {}" and continue.'.format(inps.dateList[0]))
# get deformation model from parsers
model, num_param = read_inps2model(inps)
## output preparation
# time_func_param: attributes
atrV = dict(atr)
atrV['FILE_TYPE'] = 'velocity'
atrV['UNIT'] = 'm/year'
atrV['START_DATE'] = inps.dateList[0]
atrV['END_DATE'] = inps.dateList[-1]
atrV['DATE12'] = '{}_{}'.format(inps.dateList[0], inps.dateList[-1])
if inps.ref_yx:
atrV['REF_Y'] = inps.ref_yx[0]
atrV['REF_X'] = inps.ref_yx[1]
if inps.ref_date:
atrV['REF_DATE'] = inps.ref_date
# time_func_param: config parameter
print('add/update the following configuration metadata:\n{}'.format(
configKeys))
for key in configKeys:
atrV[key_prefix + key] = str(vars(inps)[key])
# time_func_param: instantiate output file
ds_name_dict, ds_unit_dict = model2hdf5_dataset(model,
ds_shape=(length,
width))[1:]
writefile.layout_hdf5(inps.outfile,
metadata=atrV,
ds_name_dict=ds_name_dict,
ds_unit_dict=ds_unit_dict)
# timeseries_res: attributes + instantiate output file
if inps.save_res:
atrR = dict(atr)
for key in ['REF_DATE']:
if key in atrR.keys():
atrR.pop(key)
writefile.layout_hdf5(inps.res_file,
metadata=atrR,
ref_file=inps.timeseries_file)
## estimation
# calc number of box based on memory limit
memoryAll = (num_date + num_param * 2 + 2) * length * width * 4
if inps.bootstrap:
memoryAll += inps.bootstrapCount * num_param * length * width * 4
num_box = int(np.ceil(memoryAll * 3 / (inps.maxMemory * 1024**3)))
box_list = cluster.split_box2sub_boxes(box=(0, 0, width, length),
num_split=num_box,
dimension='y',
print_msg=True)
# loop for block-by-block IO
for i, box in enumerate(box_list):
box_wid = box[2] - box[0]
box_len = box[3] - box[1]
num_pixel = box_len * box_wid
if num_box > 1:
print('\n------- processing patch {} out of {} --------------'.
format(i + 1, num_box))
print('box width: {}'.format(box_wid))
print('box length: {}'.format(box_len))
# initiate output
m = np.zeros((num_param, num_pixel), dtype=dataType)
m_std = np.zeros((num_param, num_pixel), dtype=dataType)
# read input
print('reading data from file {} ...'.format(inps.timeseries_file))
ts_data = readfile.read(inps.timeseries_file, box=box)[0]
# referencing in time and space
# for file w/o reference info. e.g. ERA5.h5
if inps.ref_date:
print('referecing to date: {}'.format(inps.ref_date))
ref_ind = inps.dateList.index(inps.ref_date)
ts_data -= np.tile(ts_data[ref_ind, :, :],
(ts_data.shape[0], 1, 1))
if inps.ref_yx:
print('referencing to point (y, x): ({}, {})'.format(
inps.ref_yx[0], inps.ref_yx[1]))
ref_box = (inps.ref_yx[1], inps.ref_yx[0], inps.ref_yx[1] + 1,
inps.ref_yx[0] + 1)
ref_val = readfile.read(inps.timeseries_file, box=ref_box)[0]
ts_data -= np.tile(ref_val.reshape(ts_data.shape[0], 1, 1),
(1, ts_data.shape[1], ts_data.shape[2]))
ts_data = ts_data[inps.dropDate, :, :].reshape(inps.numDate, -1)
if atrV['UNIT'] == 'mm':
ts_data *= 1. / 1000.
ts_std = None
if inps.ts_std_file:
ts_std = readfile.read(inps.ts_std_file, box=box)[0]
ts_std = ts_std[inps.dropDate, :, :].reshape(inps.numDate, -1)
# set zero value to a fixed small value to avoid divide by zero
epsilon = 1e-5
ts_std[ts_std < epsilon] = epsilon
# mask invalid pixels
print('skip pixels with zero/nan value in all acquisitions')
ts_stack = np.nanmean(ts_data, axis=0)
mask = np.multiply(~np.isnan(ts_stack), ts_stack != 0.)
del ts_stack
if ts_std is not None:
print('skip pxiels with nan STD value in any acquisition')
num_std_nan = np.sum(np.isnan(ts_std), axis=0)
mask *= num_std_nan == 0
del num_std_nan
ts_data = ts_data[:, mask]
num_pixel2inv = int(np.sum(mask))
idx_pixel2inv = np.where(mask)[0]
print('number of pixels to invert: {} out of {} ({:.1f}%)'.format(
num_pixel2inv, num_pixel, num_pixel2inv / num_pixel * 100))
# go to next if no valid pixel found
if num_pixel2inv == 0:
continue
### estimation / solve Gm = d
print('estimating time functions via linalg.lstsq ...')
if inps.bootstrap:
## option 1 - least squares with bootstrapping
# Bootstrapping is a resampling method which can be used to estimate properties
# of an estimator. The method relies on independently sampling the data set with
# replacement.
print(
'estimating time function STD with bootstrap resampling ({} times) ...'
.format(inps.bootstrapCount))
# calc model of all bootstrap sampling
rng = np.random.default_rng()
m_boot = np.zeros((inps.bootstrapCount, num_param, num_pixel2inv),
dtype=dataType)
prog_bar = ptime.progressBar(maxValue=inps.bootstrapCount)
for i in range(inps.bootstrapCount):
# bootstrap resampling
boot_ind = rng.choice(inps.numDate,
size=inps.numDate,
replace=True)
boot_ind.sort()
# estimation
m_boot[i] = time_func.estimate_time_func(
model=model,
date_list=dates[boot_ind].tolist(),
dis_ts=ts_data[boot_ind],
seconds=seconds)[1]
prog_bar.update(i + 1,
suffix='iteration {} / {}'.format(
i + 1, inps.bootstrapCount))
prog_bar.close()
#del ts_data
# get mean/std among all bootstrap sampling
m[:, mask] = m_boot.mean(axis=0).reshape(num_param, -1)
m_std[:, mask] = m_boot.std(axis=0).reshape(num_param, -1)
del m_boot
else:
## option 2 - least squares with uncertainty propagation
G, m[:, mask], e2 = time_func.estimate_time_func(
model=model,
date_list=inps.dateList,
dis_ts=ts_data,
seconds=seconds)
#del ts_data
## Compute the covariance matrix for model parameters: Gm = d
# C_m_hat = (G.T * C_d^-1, * G)^-1 # linear propagation from the TS covariance matrix. (option 2.1)
# = sigma^2 * (G.T * G)^-1 # assuming obs errors are normally dist. in time. (option 2.2a)
# Based on the law of integrated expectation, we estimate the obs sigma^2 using
# the OLS estimation residual e_hat_i = d_i - d_hat_i
# sigma^2 = sigma_hat^2 * N / (N - P) (option 2.2b)
# = (e_hat.T * e_hat) / (N - P) # sigma_hat^2 = (e_hat.T * e_hat) / N
if ts_std is not None:
# option 2.1 - linear propagation from time-series covariance matrix
print(
'estimating time function STD from time-series STD pixel-by-pixel ...'
)
prog_bar = ptime.progressBar(maxValue=num_pixel2inv)
for i in range(num_pixel2inv):
idx = idx_pixel2inv[i]
try:
C_ts_inv = np.diag(1. /
np.square(ts_std[:, idx].flatten()))
m_var = np.diag(linalg.inv(
G.T.dot(C_ts_inv).dot(G))).astype(np.float32)
m_std[:, idx] = np.sqrt(m_var)
except linalg.LinAlgError:
m_std[:, idx] = np.nan
prog_bar.update(i + 1,
every=200,
suffix='{}/{} pixels'.format(
i + 1, num_pixel2inv))
prog_bar.close()
else:
# option 2.2a - assume obs errors following normal dist. in time
print(
'estimating time function STD from time-series fitting residual ...'
)
G_inv = linalg.inv(np.dot(G.T, G))
m_var = e2.reshape(1, -1) / (num_date - num_param)
m_std[:, mask] = np.sqrt(
np.dot(np.diag(G_inv).reshape(-1, 1), m_var))
# option 2.2b - simplified form for linear velocity (without matrix linear algebra)
# The STD can also be calculated using Eq. (10) from Fattahi and Amelung (2015, JGR)
# ts_diff = ts_data - np.dot(G, m)
# t_diff = G[:, 1] - np.mean(G[:, 1])
# vel_std = np.sqrt(np.sum(ts_diff ** 2, axis=0) / np.sum(t_diff ** 2) / (num_date - 2))
# write - time func params
block = [box[1], box[3], box[0], box[2]]
ds_dict = model2hdf5_dataset(model, m, m_std, mask=mask)[0]
for ds_name, data in ds_dict.items():
writefile.write_hdf5_block(inps.outfile,
data=data.reshape(box_len, box_wid),
datasetName=ds_name,
block=block)
# write - residual file
if inps.save_res:
block = [0, num_date, box[1], box[3], box[0], box[2]]
ts_res = np.ones(
(num_date, box_len * box_wid), dtype=np.float32) * np.nan
ts_res[:, mask] = ts_data - np.dot(G, m)[:, mask]
writefile.write_hdf5_block(inps.res_file,
data=ts_res.reshape(
num_date, box_len, box_wid),
datasetName='timeseries',
block=block)
return inps.outfile
def change_timeseries_ref_date(ts_file, ref_date, outfile=None, max_memory=4.0, force=False):
"""Change input file reference date to a different one.
Parameters: ts_file : str, timeseries file to be changed
ref_date : str, date in YYYYMMDD format
outfile : if str, save to a different file
if None, modify the data value in the existing input file
"""
ts_file = os.path.abspath(ts_file)
if not outfile:
outfile = ts_file
outfile = os.path.abspath(outfile)
print('-'*50)
print('change reference date for file: {}'.format(ts_file))
atr = readfile.read_attribute(ts_file)
dsName = atr['FILE_TYPE']
# if the input reference date is the same as the existing one.
if ref_date == atr.get('REF_DATE', None) and not force:
print('input refDate is the same as the existing REF_DATE.')
if outfile == ts_file:
print('Nothing to be done.')
return ts_file
else:
print('Copy {} to {}'.format(ts_file, outfile))
shutil.copy2(ts_file, outfile)
return outfile
# basic info
obj = timeseries(ts_file)
obj.open(print_msg=False)
num_date = obj.numDate
length = obj.length
width = obj.width
ref_idx = obj.dateList.index(ref_date)
# get list of boxes for block-by-block IO
num_box = int(np.ceil((num_date * length * width * 4 * 2) / (max_memory * 1024**3)))
box_list = split_box2sub_boxes(box=(0, 0, width, length),
num_split=num_box,
dimension='y',
print_msg=True)
# updating existing file or write new file
if outfile == ts_file:
mode = 'r+'
else:
mode = 'a'
# instantiate output file
writefile.layout_hdf5(outfile, ref_file=ts_file)
# loop for block-by-block IO
for i, box in enumerate(box_list):
box_width = box[2] - box[0]
box_length = box[3] - box[1]
if num_box > 1:
print('\n------- processing patch {} out of {} --------------'.format(i+1, num_box))
print('box width: {}'.format(box_width))
print('box length: {}'.format(box_length))
# reading
print('reading data ...')
ts_data = readfile.read(ts_file, box=box)[0]
print('referencing in time ...')
dshape = ts_data.shape
ts_data -= np.tile(ts_data[ref_idx, :, :].reshape(1, dshape[1], dshape[2]), (dshape[0], 1, 1))
# writing
block = (0, num_date, box[1], box[3], box[0], box[2])
writefile.write_hdf5_block(outfile,
data=ts_data,
datasetName=dsName,
block=block,
mode=mode)
# update metadata
print('update "REF_DATE" attribute value to {}'.format(ref_date))
with h5py.File(outfile, 'r+') as f:
f.attrs['REF_DATE'] = ref_date
f.attrs['FILE_PATH'] = outfile
return outfile
def subset_file(fname, subset_dict_input, out_file=None):
"""Subset file with
Inputs:
fname : str, path/name of file
out_file : str, path/name of output file
subset_dict : dict, subsut parameter, including the following items:
subset_x : list of 2 int, subset in x direction, default=None
subset_y : list of 2 int, subset in y direction, default=None
subset_lat : list of 2 float, subset in lat direction, default=None
subset_lon : list of 2 float, subset in lon direction, default=None
fill_value : float, optional. filled value for area outside of data coverage. default=None
None/not-existed to subset within data coverage only.
tight : bool, tight subset or not, for lookup table file, i.e. geomap*.trans
Outputs:
out_file : str, path/name of output file;
out_file = 'subset_'+fname, if fname is in current directory;
out_file = fname, if fname is not in the current directory.
"""
# Input File Info
atr = readfile.read_attribute(fname)
width = int(atr['WIDTH'])
length = int(atr['LENGTH'])
k = atr['FILE_TYPE']
print('subset ' + k + ' file: ' + fname + ' ...')
subset_dict = subset_dict_input.copy()
# Read Subset Inputs into 4-tuple box in pixel and geo coord
pix_box, geo_box = subset_input_dict2box(subset_dict, atr)
coord = ut.coordinate(atr)
# if fill_value exists and not None, subset data and fill assigned value for area out of its coverage.
# otherwise, re-check subset to make sure it's within data coverage and initialize the matrix with np.nan
outfill = False
if 'fill_value' in subset_dict.keys() and subset_dict['fill_value']:
outfill = True
else:
outfill = False
if not outfill:
pix_box = coord.check_box_within_data_coverage(pix_box)
subset_dict['fill_value'] = np.nan
geo_box = coord.box_pixel2geo(pix_box)
data_box = (0, 0, width, length)
print('data range in (x0,y0,x1,y1): {}'.format(data_box))
print('subset range in (x0,y0,x1,y1): {}'.format(pix_box))
print('data range in (W, N, E, S): {}'.format(
coord.box_pixel2geo(data_box)))
print('subset range in (W, N, E, S): {}'.format(geo_box))
if pix_box == data_box:
print('Subset range == data coverage, no need to subset. Skip.')
return fname
# Calculate Subset/Overlap Index
pix_box4data, pix_box4subset = get_box_overlap_index(data_box, pix_box)
########################### Data Read and Write ######################
# Output File Name
if not out_file:
if os.getcwd() == os.path.dirname(os.path.abspath(fname)):
if 'tight' in subset_dict.keys() and subset_dict['tight']:
out_file = '{}_tight{}'.format(
os.path.splitext(fname)[0],
os.path.splitext(fname)[1])
else:
out_file = 'sub_' + os.path.basename(fname)
else:
out_file = os.path.basename(fname)
print('writing >>> ' + out_file)
# update metadata
atr = attr.update_attribute4subset(atr, pix_box)
# subset datasets one by one
dsNames = readfile.get_dataset_list(fname)
maxDigit = max([len(i) for i in dsNames])
ext = os.path.splitext(out_file)[1]
if ext in ['.h5', '.he5']:
# initiate the output file
writefile.layout_hdf5(out_file, metadata=atr, ref_file=fname)
# subset dataset one-by-one
for dsName in dsNames:
with h5py.File(fname, 'r') as fi:
ds = fi[dsName]
ds_shape = ds.shape
ds_ndim = ds.ndim
print('cropping {d} in {b} from {f} ...'.format(
d=dsName, b=pix_box4data, f=os.path.basename(fname)))
if ds_ndim == 2:
# read
data = ds[pix_box4data[1]:pix_box4data[3],
pix_box4data[0]:pix_box4data[2]]
# crop
data_out = np.ones(
(pix_box[3] - pix_box[1], pix_box[2] - pix_box[0]),
data.dtype) * subset_dict['fill_value']
data_out[pix_box4subset[1]:pix_box4subset[3],
pix_box4subset[0]:pix_box4subset[2]] = data
data_out = np.array(data_out, dtype=data.dtype)
# write
block = [0, int(atr['LENGTH']), 0, int(atr['WIDTH'])]
writefile.write_hdf5_block(out_file,
data=data_out,
datasetName=dsName,
block=block,
print_msg=True)
if ds_ndim == 3:
prog_bar = ptime.progressBar(maxValue=ds_shape[0])
for i in range(ds_shape[0]):
# read
data = ds[i, pix_box4data[1]:pix_box4data[3],
pix_box4data[0]:pix_box4data[2]]
# crop
data_out = np.ones(
(1, pix_box[3] - pix_box[1],
pix_box[2] - pix_box[0]),
data.dtype) * subset_dict['fill_value']
data_out[:, pix_box4subset[1]:pix_box4subset[3],
pix_box4subset[0]:pix_box4subset[2]] = data
# write
block = [
i, i + 1, 0,
int(atr['LENGTH']), 0,
int(atr['WIDTH'])
]
writefile.write_hdf5_block(out_file,
data=data_out,
datasetName=dsName,
block=block,
print_msg=False)
prog_bar.update(i + 1,
suffix='{}/{}'.format(
i + 1, ds_shape[0]))
prog_bar.close()
print('finished writing to file: {}'.format(out_file))
else:
# IO for binary files
dsDict = dict()
for dsName in dsNames:
dsDict[dsName] = subset_dataset(
fname,
dsName,
pix_box,
pix_box4data,
pix_box4subset,
fill_value=subset_dict['fill_value'])
writefile.write(dsDict,
out_file=out_file,
metadata=atr,
ref_file=fname)
# write extra metadata files for ISCE data files
if os.path.isfile(fname + '.xml') or os.path.isfile(fname +
'.aux.xml'):
# write ISCE XML file
dtype_gdal = readfile.NUMPY2GDAL_DATATYPE[atr['DATA_TYPE']]
dtype_isce = readfile.GDAL2ISCE_DATATYPE[dtype_gdal]
writefile.write_isce_xml(out_file,
width=int(atr['WIDTH']),
length=int(atr['LENGTH']),
bands=len(dsDict.keys()),
data_type=dtype_isce,
scheme=atr['scheme'],
image_type=atr['FILE_TYPE'])
print(f'write file: {out_file}.xml')
# write GDAL VRT file
if os.path.isfile(fname + '.vrt'):
from isceobj.Util.ImageUtil import ImageLib as IML
img = IML.loadImage(out_file)[0]
img.renderVRT()
print(f'write file: {out_file}.vrt')
return out_file
def multilook_file(infile,
lks_y,
lks_x,
outfile=None,
method='average',
margin=[0, 0, 0, 0],
max_memory=4):
""" Multilook input file
Parameters: infile - str, path of input file to be multilooked.
lks_y - int, number of looks in y / row direction.
lks_x - int, number of looks in x / column direction.
margin - list of 4 int, number of pixels to be skipped during multilooking.
useful for offset product, where the marginal pixels are ignored during
cross correlation matching.
outfile - str, path of output file
Returns: outfile - str, path of output file
"""
lks_y = int(lks_y)
lks_x = int(lks_x)
# input file info
atr = readfile.read_attribute(infile)
length, width = int(atr['LENGTH']), int(atr['WIDTH'])
k = atr['FILE_TYPE']
print('multilooking {} {} file: {}'.format(atr['PROCESSOR'], k, infile))
print('number of looks in y / azimuth direction: %d' % lks_y)
print('number of looks in x / range direction: %d' % lks_x)
print('multilook method: {}'.format(method))
# margin --> box
if margin is not [0, 0, 0, 0]: # top, bottom, left, right
box = (margin[2], margin[0], width - margin[3], length - margin[1])
print(
'number of pixels to skip in top/bottom/left/right boundaries: {}'.
format(margin))
else:
box = (0, 0, width, length)
# output file name
ext = os.path.splitext(infile)[1]
if not outfile:
if os.getcwd() == os.path.dirname(os.path.abspath(infile)):
outfile = os.path.splitext(infile)[0] + '_' + str(
lks_y) + 'alks_' + str(lks_x) + 'rlks' + ext
else:
outfile = os.path.basename(infile)
# update metadata
atr = attr.update_attribute4multilook(atr, lks_y, lks_x, box=box)
if ext in ['.h5', '.he5']:
writefile.layout_hdf5(outfile, metadata=atr, ref_file=infile)
# read source data and multilooking
dsNames = readfile.get_dataset_list(infile)
maxDigit = max([len(i) for i in dsNames])
dsDict = dict()
for dsName in dsNames:
print('multilooking {d:<{w}} from {f} ...'.format(
d=dsName, w=maxDigit, f=os.path.basename(infile)))
# split in Y/row direction for IO for HDF5 only
if ext in ['.h5', '.he5']:
# calc step size with memory usage up to 4 GB
with h5py.File(infile, 'r') as f:
ds = f[dsName]
ds_size = np.prod(ds.shape) * 4
num_step = int(np.ceil(ds_size * 4 / (max_memory * 1024**3)))
row_step = int(np.rint(length / num_step / 10) * 10)
row_step = max(row_step, 10)
else:
row_step = box[3] - box[1]
num_step = int(np.ceil((box[3] - box[1]) / (row_step * lks_y)))
for i in range(num_step):
r0 = box[1] + row_step * lks_y * i
r1 = box[1] + row_step * lks_y * (i + 1)
r1 = min(r1, box[3])
# IO box
box_i = (box[0], r0, box[2], r1)
box_o = (int((box[0] - box[0]) / lks_x), int(
(r0 - box[1]) / lks_y), int(
(box[2] - box[0]) / lks_x), int((r1 - box[1]) / lks_y))
print('box: {}'.format(box_o))
# read / multilook
if method == 'nearest':
data = readfile.read(infile,
datasetName=dsName,
box=box_i,
xstep=lks_x,
ystep=lks_y,
print_msg=False)[0]
else:
data = readfile.read(infile,
datasetName=dsName,
box=box_i,
print_msg=False)[0]
data = multilook_data(data, lks_y, lks_x)
# output block
if data.ndim == 3:
block = [
0, data.shape[0], box_o[1], box_o[3], box_o[0], box_o[2]
]
else:
block = [box_o[1], box_o[3], box_o[0], box_o[2]]
# write
if ext in ['.h5', '.he5']:
writefile.write_hdf5_block(outfile,
data=data,
datasetName=dsName,
block=block,
print_msg=False)
else:
dsDict[dsName] = data
# for binary file with 2 bands, always use BIL scheme
if (len(dsDict.keys()) == 2
and os.path.splitext(infile)[1] not in ['.h5', '.he5']
and atr.get('scheme', 'BIL').upper() != 'BIL'):
print('the input binary file has 2 bands with band interleave as: {}'.
format(atr['scheme']))
print(
'for the output binary file, change the band interleave to BIL as default.'
)
atr['scheme'] = 'BIL'
if ext not in ['.h5', '.he5']:
writefile.write(dsDict,
out_file=outfile,
metadata=atr,
ref_file=infile)
# write extra metadata files for ISCE data files
if os.path.isfile(infile + '.xml') or os.path.isfile(infile +
'.aux.xml'):
# write ISCE XML file
dtype_gdal = readfile.NUMPY2GDAL_DATATYPE[atr['DATA_TYPE']]
dtype_isce = readfile.GDAL2ISCE_DATATYPE[dtype_gdal]
writefile.write_isce_xml(outfile,
width=int(atr['WIDTH']),
length=int(atr['LENGTH']),
bands=len(dsDict.keys()),
data_type=dtype_isce,
scheme=atr['scheme'],
image_type=atr['FILE_TYPE'])
print(f'write file: {outfile}.xml')
# write GDAL VRT file
if os.path.isfile(infile + '.vrt'):
from isceobj.Util.ImageUtil import ImageLib as IML
img = IML.loadImage(outfile)[0]
img.renderVRT()
print(f'write file: {outfile}.vrt')
return outfile
