def getElementPath(element):
""" return a qualified name for the element if it is possible to compute one
by concatenating the "name" of parent elements.
The path is computed according to some appropriate
"parentship" association depending on the type of elements.
If it is not possible to get the path, then return the id of the element.
"""
try:
names = map( lambda x:x.getName(),
getElementParents(element,inclusive=True,reverse=True))
if exists(isEmpty,names):
return unicode(getElementId(element))
else:
return unicode(".".join(names))
except:
return unicode(getElementId(element))
import os
import sys
from misc import args, sep, exists, pd, run, err
if len(args) < 4:
err("python3 raster_stack.py [input raster 1] [input raster 2] .." +
" [input raster n] [output raster name]")
rasters = args[1:-1] # raster files to cat together
outf = args[-1] # output filename
ofhn = args[-1][:-4] + '.hdr' # output header file name
cmd = ['cat'] + rasters + ['>', outf] # cat command
cmd = ' '.join(cmd)
print(cmd)
print('')
if not exists(outf):
a = os.system(cmd)
for r in rasters:
print(" ", r)
cmd = [
'python3', # envi_header_cat.py is almost like a reverse-polish notation. Have to put the "first thing" on the back..
pd + 'envi_header_cat.py',
rasters[1][:-4] + '.hdr',
rasters[0][:-4] + '.hdr',
ofhn
] # 'raster.hdr']
cmd = ' '.join(cmd)
run(cmd)
fn, s = args[1], args[2]
cd = pd + '..' + sep + 'cpp' + sep
hfn = hdr_fn(fn)
samples, lines, bands = read_hdr(hfn)
nrow, ncol = lines, samples
bi = []
bn = band_names(hfn)
for i in range(len(bn)):
if len(bn[i].split(s)) > 1:
bi.append(i)
print(' ', str(i + 1), ' ', bn[i])
if not exists(fn + '_001.hdr'):
run(cd + 'unstack.exe ' + fn)
for i in bi:
f = fn + '_' + str(i + 1).zfill(3) + '.bin'
print(f)
fni = [fn + '_' + str(i + 1).zfill(3) + '.bin' for i in bi]
bands = str(len(bi)) # number of bands to write
ofn = fn + '_' + s + '.bin'
ohn = fn + '_' + s + '.hdr'
run('cp ' + hfn + ' ' + ohn)
c = ['python3', pd + 'envi_header_modify.py', ohn, nrow, ncol, bands]
c += [('"' + bn[bi[i]] + '"') for i in range(len(bi))]
c = ' '.join(c)
run(c)
import sys
from misc import err, run, pd, sep, exists, args
cd = pd + '..' + sep + 'cpp' + sep
if len(args) < 2:
err('sentinel2_trace_active.py [input file, binary mask envi type 4] [optional arg: class index]'
)
class_select = None
if len(args) > 2:
class_select = int(args[2])
print(class_select)
fn = args[1]
if not exists(fn):
err('please check input file')
for i in [150]: # [10, 20, 60]: # 90
if not exists(fn + '_flood4.bin'):
run('ulimit -s 1000000')
run(cd + 'flood.exe ' + fn)
if not exists(fn + '_flood4.bin_link.bin'):
run(cd + 'class_link.exe ' + fn + '_flood4.bin ' + str(i)) # 40')
if not exists(fn + '_flood4.bin_link.bin_recode.bin'):
run(cd + 'class_recode.exe ' + fn + '_flood4.bin_link.bin 1')
if not exists(fn + '_flood4.bin_link.bin_recode.bin_wheel.bin'):
run(cd + 'class_wheel.exe ' + fn + '_flood4.bin_link.bin_recode.bin')
run('python3 ' + pd + 'raster_plot.py ' + fn +
'_flood4.bin_link.bin_recode.bin_wheel.bin 1 2 3 1 1')
print('class onehot..')
# 2) run the fire detection!
cmds = []
dets = []
for x in L:
# print(x)
df = 'SENTINEL2_L2A_EPSG*10m.bin'
cmd = 'find ' + x[1] + ' -name ' + df
y = os.popen(cmd).read().strip()
if y == '':
print('x warning: no data ' + x[1])
else:
# run detector
fn = y + '_active.bin'
print('* ' + fn)
dets.append([x[0], x[1], y, fn])
if not exists(fn):
cmd = pd + '../cpp/sentinel2_active.exe ' + y
print(cmd)
cmds.append(cmd)
def r(x):
return os.popen(x).read()
parfor(r, cmds, 4) # limited by disk I/O
# 3) cat the results together?
nrow, ncol, nband = 0, 0, 0
cmd = 'cat '
for d in dets:
def make_hist_se(infile, outfile, hashing_val, l10p_cutoff, l2fc_cutoff, all_exons, exon_overhang, intron_overhang):
try:
region_types = ["upstream_region_skipped_exon",
"upstream_region_downstream_exon",
"downstream_region_skipped_exon",
"downstream_region_upstream_exon"]
position_sum = {}
count = []
event_dict = defaultdict(dict)
event_dict['downstream_region_upstream_exon'] = {}
event_dict['upstream_region_skipped_exon'] = {}
event_dict['downstream_region_skipped_exon'] = {}
event_dict['upstream_region_downstream_exon'] = {}
intersect = 0
no_intersect = 0
with open(infile, 'r') as f:
for line in f:
intersects_any_region = False # if this flag is true, then a peak intersects a region.
line = line.split('\t')
chrom = line[0]
pstart = int(line[1])
pstop = int(line[2])
l10p = float(line[3])
l2fc = float(line[4])
stra = line[5].strip()
# correct bed files being 0-based, open ended
pstart = pstart + 1
if l10p < l10p_cutoff:
continue
if l2fc < l2fc_cutoff:
continue
x = int(pstart / hashing_val)
y = int(pstop / hashing_val)
p_str = '{}:{}-{}:{}'.format(chrom, pstart, pstop, stra)
# for each peak, find ALL regions that intersect it
for region_type in region_types: # within a region
tmphash = {}
for i in range(x, y + 1): # within a bin
for event in all_exons[chrom, stra, i, region_type]:
exchr, exregstart, exstart, exregstop, exstr = event.split(':')
if pstop < int(exregstart):
continue
if pstart > int(exregstop): # pass if peak stop occurs before exon region start
continue
tmphash[event] = 1 # otherwise peak falls within event region
intersects_any_region = True
# event_dict[region_type][p_str] = tmphash
for event in tmphash:
if stra == "+":
exchr, exregstart, exstart, exregstop, exstr = event.split(':')
start_val = max(int(pstart), int(exregstart)) # peak start OR region start
end_val = min(int(pstop), int(exregstop)) # peak stop OR region stop
for j in range(start_val,
end_val + 1): # count intersecting positions between peak and region
relative_pos = j - int(exstart) # calculate relative position
position_sum[region_type, relative_pos] = misc.ini(position_sum,
region_type,
relative_pos) # count + 1 for the region
elif stra == '-':
exchr, exregstart, exstart, exregstop, exstr = event.split(':')
start_val = max(int(pstart), int(exregstart))
end_val = min(int(pstop), int(exregstop))
for j in range(start_val, end_val + 1):
relative_pos = -1 * (j - int(exstart))
position_sum[region_type, relative_pos] = misc.ini(position_sum,
region_type,
relative_pos)
else:
print("strand error\n")
if intersects_any_region:
intersect+=1
else:
no_intersect+=1
o = open(outfile + '.overlapping_peaks', 'w')
o.write('infile\tintersect\tno_intersect\n')
o.write('{}\t{}\t{}'.format(infile, intersect, no_intersect))
o.close()
# print(event_dict['downstream_region_skipped_exon']['chr11:47237411-47237505:+'])
# count from 0 to max
current_pos = 0
o = open(outfile, 'w')
for j in range(-exon_overhang, intron_overhang + 1):
if misc.exists(position_sum, "downstream_region_upstream_exon", j):
o.write("{}\n".format(position_sum["downstream_region_upstream_exon", j]))
count.append(position_sum["downstream_region_upstream_exon", j])
else:
o.write("{}\n".format(0))
count.append(0)
current_pos = current_pos + 1
for j in range(-intron_overhang, exon_overhang + 1):
if misc.exists(position_sum, "upstream_region_skipped_exon", j):
o.write("{}\n".format(position_sum["upstream_region_skipped_exon", j]))
count.append(position_sum["upstream_region_skipped_exon", j])
else:
o.write("{}\n".format(0))
count.append(0)
current_pos = current_pos + 1
for j in range(-exon_overhang, intron_overhang + 1):
if misc.exists(position_sum, "downstream_region_skipped_exon", j):
o.write("{}\n".format(position_sum["downstream_region_skipped_exon", j]))
count.append(position_sum["downstream_region_skipped_exon", j])
else:
o.write("{}\n".format(0))
count.append(0)
current_pos = current_pos + 1
for j in range(-intron_overhang, exon_overhang + 1):
if misc.exists(position_sum, "upstream_region_downstream_exon", j):
o.write("{}\n".format(position_sum["upstream_region_downstream_exon", j]))
count.append(position_sum["upstream_region_downstream_exon", j])
else:
o.write("{}\n".format(0))
count.append(0)
current_pos = current_pos + 1
o.close()
return count
except Exception as e:
print(e)
def make_hist_a5ss(infile, outfile, hashing_val, l10p_cutoff, l2fc_cutoff, all_exons, exon_overhang, intron_overhang):
try:
# region_types = ["ex_up","dnex","ex_dn","upex"] # former annotation
region_types = ["upstream_region_skipped_exon",
"downstream_region_skipped_exon",
"upstream_region_downstream_exon",
"downstream_region_upstream_exon"]
position_sum = {}
count = 0
with open(infile, 'r') as f:
for line in f:
line = line.split('\t')
chrom = line[0]
pstart = int(line[1])
pstop = int(line[2])
l10p = float(line[3])
l2fc = float(line[4])
stra = line[5].strip()
# correct bed files being 0-based, open ended
pstart = pstart + 1
if l10p < l10p_cutoff:
continue
if l2fc < l2fc_cutoff:
continue
x = int(pstart / hashing_val)
y = int(pstop / hashing_val)
# for each peak, find ALL regions that intersect it
for region_type in region_types: # within a region
tmphash = {}
for i in range(x, y + 1): # within a bin
for event in all_exons[chrom, stra, i, region_type]:
exchr, exregstart, exstart, exregstop, exstr = event.split(':')
if pstop < int(exregstart): # pass if peak stop occurs before exon region start
continue
if pstart > int(exregstop): # pass if peak start occurs after exon region end
continue
tmphash[event] = 1 # otherwise peak falls within event region
for event in tmphash:
if stra == "+":
exchr, exregstart, exstart, exregstop, exstr = event.split(':')
start_val = max(int(pstart), int(exregstart)) # peak start OR region start
end_val = min(int(pstop), int(exregstop)) # peak stop OR region stop
for j in range(start_val,
end_val + 1): # count intersecting positions between peak and region
relative_pos = j - int(exstart) # calculate relative position
position_sum[region_type, relative_pos] = misc.ini(position_sum,
region_type,
relative_pos) # count + 1 for the region
elif stra == '-':
exchr, exregstart, exstart, exregstop, exstr = event.split(':')
start_val = max(int(pstart), int(exregstart))
end_val = min(int(pstop), int(exregstop))
for j in range(start_val, end_val + 1):
relative_pos = -1 * (j - int(exstart))
position_sum[region_type, relative_pos] = misc.ini(position_sum,
region_type,
relative_pos)
else:
print("strand error\n")
# count from 0 to max
current_pos = 0
o = open(outfile, 'w')
for j in range(-exon_overhang, 1):
if misc.exists(position_sum, "downstream_region_upstream_exon", j):
o.write("{}\n".format(position_sum["downstream_region_upstream_exon", j]))
else:
o.write("{}\n".format(0))
current_pos = current_pos + 1
# o.write("STOP")
for j in range(0, exon_overhang + 1):
if misc.exists(position_sum, "upstream_region_skipped_exon", j):
o.write("{}\n".format(position_sum["upstream_region_skipped_exon", j]))
else:
o.write("{}\n".format(0))
current_pos = current_pos + 1
# o.write("STOP")
for j in range(-exon_overhang, intron_overhang + 1):
if misc.exists(position_sum, "downstream_region_skipped_exon", j):
o.write("{}\n".format(position_sum["downstream_region_skipped_exon", j]))
else:
o.write("{}\n".format(0))
current_pos = current_pos + 1
for j in range(-intron_overhang, exon_overhang + 1):
if misc.exists(position_sum, "upstream_region_downstream_exon", j):
o.write("{}\n".format(position_sum["upstream_region_downstream_exon", j]))
else:
o.write("{}\n".format(0))
current_pos = current_pos + 1
o.close()
except Exception as e:
print(e)
