-
Notifications
You must be signed in to change notification settings - Fork 3
/
zonal_stats_by_raster.py
205 lines (186 loc) · 8.47 KB
/
zonal_stats_by_raster.py
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
61
62
63
64
65
66
67
68
69
70
71
72
73
74
75
76
77
78
79
80
81
82
83
84
85
86
87
88
89
90
91
92
93
94
95
96
97
98
99
100
101
102
103
104
105
106
107
108
109
110
111
112
113
114
115
116
117
118
119
120
121
122
123
124
125
126
127
128
129
130
131
132
133
134
135
136
137
138
139
140
141
142
143
144
145
146
147
148
149
150
151
152
153
154
155
156
157
158
159
160
161
162
163
164
165
166
167
168
169
170
171
172
173
174
175
176
177
178
179
180
181
182
183
184
185
186
187
188
189
190
191
192
193
194
195
196
197
198
199
200
201
202
203
204
205
"""Calculate stats per landcover code type."""
import argparse
import os
import logging
import hashlib
import multiprocessing
import time
from osgeo import gdal
from ecoshard import geoprocessing
from ecoshard import taskgraph
import numpy
gdal.SetCacheMax(2**26)
_LARGEST_BLOCK = 2**26
_GTIFF_CREATION_TUPLE_OPTIONS = ('GTIFF', (
geoprocessing.DEFAULT_GTIFF_CREATION_TUPLE_OPTIONS[1]) +
('SPARSE_OK=TRUE',))
logging.basicConfig(
level=logging.DEBUG,
format=(
'%(asctime)s (%(relativeCreated)d) %(levelname)s %(name)s'
' [%(pathname)s.%(funcName)s:%(lineno)d] %(message)s'))
LOGGER = logging.getLogger(__name__)
logging.getLogger('ecoshard.taskgraph').setLevel(logging.INFO)
def mask_out_op(mask_data, base_data, mask_code, base_nodata):
"""Return 1 where base data == mask_code, 0 or nodata othewise."""
result = numpy.full(base_data.shape, base_nodata, dtype=numpy.float32)
valid_mask = (mask_data == mask_code) & (~numpy.isnan(base_data))
if numpy.any(valid_mask):
result[valid_mask] = base_data[valid_mask]
else:
result = None
return result
def _calculate_stats(
aligned_raster_path_list, mask_code, masked_nodata,
masked_raster_path):
LOGGER.debug(f'_calculate_stats for {masked_raster_path}')
geoprocessing.raster_calculator(
[(aligned_raster_path_list[0], 1), (aligned_raster_path_list[1], 1),
(mask_code, 'raw'), (masked_nodata, 'raw')],
mask_out_op, masked_raster_path, gdal.GDT_Float32,
masked_nodata,
raster_driver_creation_tuple=_GTIFF_CREATION_TUPLE_OPTIONS)
masked_raster = gdal.OpenEx(masked_raster_path, gdal.OF_RASTER)
masked_band = masked_raster.GetRasterBand(1)
n_pixels = masked_band.XSize * masked_band.YSize
LOGGER.debug(f'{n_pixels}')
(raster_min, raster_max, raster_mean, raster_stdev) = (
masked_band.GetStatistics(0, 1))
masked_band = None
masked_raster = None
LOGGER.debug(f'getting stats for {masked_raster_path}')
valid_count = 0
for offset_dict, masked_block in geoprocessing.iterblocks(
(masked_raster_path, 1), largest_block=_LARGEST_BLOCK,
skip_sparse=True):
valid_mask = (masked_block != masked_nodata)
valid_count += numpy.count_nonzero(valid_mask)
nodata_count = n_pixels - valid_count
LOGGER.debug(f'{valid_count}')
return (raster_min, raster_max, raster_mean,
raster_stdev, valid_count, nodata_count)
if __name__ == '__main__':
parser = argparse.ArgumentParser(description='Zonal stats by raster')
parser.add_argument(
'landcover_raster', help='Path to landcover raster.')
parser.add_argument(
'other_raster', help='Path to another raster to calculate stats over.')
parser.add_argument(
'--working_dir', default='lulc_raster_stats_workspace',
help='location to store temporary files')
parser.add_argument(
'--do_not_align', default=False, action='store_true',
help='pass this flag to avoid aligning rasters')
parser.add_argument('--basename', type=str, help=(
'output table will include this name, if left off a unique hash will '
'be created from the landcover and other raster filepath and '
'timestamp strings, this means a new subworkspace will be created on '
'each run.'))
parser.add_argument(
'--n_workers', type=int, default=multiprocessing.cpu_count(),
help=(
'number of CPUs to use for processing, default is all CPUs on '
'the machine'))
parser.add_argument(
'--ndv', type=float, help=(
'set the nodata value if one is not defined or you wish to '
'override that value when calculating statistics'))
args = parser.parse_args()
if args.basename:
basename = args.basename
else:
basename = hashlib.sha1(
f'{args.landcover_raster}_{args.other_raster}'.encode(
'utf-8')).hexdigest()[:12]
basename += '_'+time.strftime("%Y_%m_%d_%H_%M_%S", time.gmtime())
working_dir = os.path.join(args.working_dir, basename)
os.makedirs(working_dir, exist_ok=True)
task_graph = taskgraph.TaskGraph(args.working_dir, args.n_workers, 15.0)
other_raster_info = geoprocessing.get_raster_info(args.other_raster)
align_hash = hashlib.md5(
f'{other_raster_info["pixel_size"]}_'
f'{other_raster_info["projection_wkt"]}'.encode('utf-8')).hexdigest()
aligned_raster_dir = os.path.join(
args.working_dir, f'aligned_rasters_{align_hash}')
os.makedirs(aligned_raster_dir, exist_ok=True)
base_raster_path_list = [args.landcover_raster, args.other_raster]
aligned_raster_path_list = [
os.path.join(aligned_raster_dir, os.path.basename(path))
for path in base_raster_path_list]
other_nodata = other_raster_info['nodata'][0]
if args.ndv is None and (
other_nodata is None or
not numpy.isfinite(other_raster_info['nodata'][0])):
raise ValueError(
f'nodata value undefined for {args.other_raster}, you must set it '
f'using the --ndv flag. -9999 is a good value if you are unsure '
f'of what to select.')
if args.ndv is not None:
other_nodata = args.ndv
if not args.do_not_align and (args.landcover_raster != args.other_raster):
# check if target pixel size is not much larger than base
landcover_raster_info = geoprocessing.get_raster_info(
args.landcover_raster)
if any([abs(other_raster_info['pixel_size'][ix]) /
abs(landcover_raster_info['pixel_size'][ix]) >= 2
for ix in [0, 1]]):
interpolation_list = ['mode', 'near']
else:
interpolation_list = ['near', 'near']
align_task = task_graph.add_task(
func=geoprocessing.align_and_resize_raster_stack,
args=(
base_raster_path_list, aligned_raster_path_list,
interpolation_list, other_raster_info['pixel_size'],
'intersection',
),
kwargs={
'target_projection_wkt': other_raster_info['projection_wkt'],
'raster_driver_creation_tuple': _GTIFF_CREATION_TUPLE_OPTIONS,
'gdal_warp_options': (
'warpMemoryLimit=1000', 'multithread=TRUE')},
target_path_list=aligned_raster_path_list,
task_name=f'aligning {aligned_raster_path_list}')
else:
aligned_raster_path_list = base_raster_path_list
align_task = task_graph.add_task()
task_graph.join()
lulc_nodata = geoprocessing.get_raster_info(
args.landcover_raster)['nodata']
LOGGER.info('calculate unique values')
unique_value_task = task_graph.add_task(
func=geoprocessing.get_unique_values,
args=((args.landcover_raster, 1),),
store_result=True,
dependent_task_list=[align_task],
task_name=f'unique values for {args.landcover_raster}')
unique_values = unique_value_task.get()
LOGGER.debug(unique_values)
stats_table = open(f'stats_table_{basename}.csv', 'w')
stats_table.write(
'lucode,min,max,mean,stdev,valid_count,nodata_count,total\n')
masked_stats_list = []
for mask_code in sorted(unique_values):
LOGGER.debug(f'scheduling {mask_code}')
masked_raster_path = os.path.join(working_dir, '%d.tif' % mask_code)
stats_task = task_graph.add_task(
func=_calculate_stats,
args=(
aligned_raster_path_list, mask_code, other_nodata,
masked_raster_path),
store_result=True,
dependent_task_list=[unique_value_task],
target_path_list=[masked_raster_path],
task_name=f'mask {masked_raster_path}')
masked_stats_list.append((stats_task, mask_code))
LOGGER.debug('waiting for it to gadot')
for masked_task, mask_code in masked_stats_list:
(raster_min, raster_max, raster_mean,
raster_stdev, valid_count, nodata_count) = masked_task.get()
stats_table.write(
'%d,%f,%f,%f,%f,%d,%d,%d\n' % (
mask_code, raster_min, raster_max, raster_mean, raster_stdev,
valid_count, nodata_count, valid_count+nodata_count))
stats_table.close()
task_graph.join()
task_graph.close()