def upload_asset_s3(self, params):
try:
filepath = params['filepath']
gaia = Gaia()
# uploading asset to s3
storage_url = gaia.store_asset_to_s3(filepath)
asset = params['asset']
asset.storage_url = storage_url
asset.save()
# Remove the local copy of data
shutil.rmtree(filepath)
print('successfully download to amazon s3')
return storage_url
except:
return 'Error On Uploading Asset TO s3'
def generate_analytic_assets(params):
'''
Third step is to do calculation, generate the asset, store it S3 and our DB
'''
gaia = Gaia()
theia = Theia()
athena = Athena()
date = params['date']
aoi_id = params['aoi_id']
filepath = params['filepath']
planet_item_id = params['planet_item_id']
aoi = Aoi.objects.get_by_id(aoi_id)
udm_filename = glob.glob(filepath + '*udm_clip.tif')[0]
raw_filename = glob.glob(filepath + '*_AnalyticMS_clip.tif')[0]
metadata = glob.glob(filepath + '*_AnalyticMS_metadata_clip.xml')[0]
udm, err = gaia.read_band(udm_filename)
metadata_xml, err = gaia.parse_xml(metadata)
band_file, err = gaia.read_band(raw_filename)
clip_udm = athena.create_unusable_clip_mask(band_file, udm)
is_hazy = athena.is_hazy(band_file, metadata_xml, clip_udm)
ndvi, err = athena.calculate_NDVI(band_file, metadata_xml)
ndwi, err = athena.calculate_NDWI(band_file, metadata_xml)
bai, err = athena.calculate_BAI(band_file, metadata_xml)
rvi, err = athena.calculate_RVI(band_file, metadata_xml)
gndvi, err = athena.calculate_GNDVI(band_file, metadata_xml)
msavi, err = athena.calculate_MSAVI(band_file, metadata_xml)
dirt, err = athena.calculate_DIRT(band_file, metadata_xml)
evi, err = athena.calculate_EVI(band_file, metadata_xml)
usability_score = athena.calculate_usability_score(udm)
ndvi = np.where(clip_udm, -1, ndvi)
ndwi = np.where(clip_udm, -1, ndwi)
bai = np.where(clip_udm, -1, bai)
rvi = np.where(clip_udm, -1, rvi)
gndvi = np.where(clip_udm, -1, gndvi)
msavi = np.where(clip_udm, -1, msavi)
dirt = np.where(clip_udm, -1, dirt)
evi = np.where(clip_udm, -1, evi)
mean_ndvi = np.ma.array(data = ndvi, mask = clip_udm).mean()
mean_ndwi = np.ma.array(data = ndwi, mask = clip_udm).mean()
mean_bai = np.ma.array(data = bai, mask = clip_udm).mean()
mean_rvi = np.ma.array(data = rvi, mask = clip_udm).mean()
mean_gndvi = np.ma.array(data = gndvi, mask = clip_udm).mean()
mean_msavi = np.ma.array(data = msavi, mask = clip_udm).mean()
mean_dirt = np.ma.array(data = dirt, mask = clip_udm).mean()
mean_evi = np.ma.array(data = evi, mask = clip_udm).mean()
print('-----')
print(date)
print("ndvi : " + str(mean_ndvi))
print("ndwi : " + str(mean_ndwi))
print("bai : " + str(mean_bai))
print("rvi : " + str(mean_rvi))
print("gndvi : " + str(mean_gndvi))
print("msavi : " + str(mean_msavi))
print("dirt : " + str(mean_dirt))
print("evi : " + str(mean_evi))
print('-----')
bands_arr = (ndvi, ndwi, bai, rvi, gndvi, msavi, dirt, evi)
file_profile = band_file.profile
file_profile['count'] = len(bands_arr)
file_profile['dtype'] = 'float64'
gaia.write_band_to_tiff({
'bands' : bands_arr,
'profile' : file_profile,
'filepath' : filepath,
'filename' : 'output-analytics.tif'
})
theia.create_cmap_asset(ndvi, filepath, 'output-cmap-ndvi')
theia.create_cmap_asset(ndwi, filepath, 'output-cmap-ndwi', cmap_name='NDWI')
theia.create_cmap_asset(bai, filepath, 'output-cmap-bai')
theia.create_cmap_asset(rvi, filepath, 'output-cmap-rvi', cmap_name='RVI', vmin=-3, vmax=3)
theia.create_cmap_asset(gndvi, filepath, 'output-cmap-gndvi')
theia.create_cmap_asset(msavi, filepath, 'output-cmap-msavi', cmap_name='MSAVI')
theia.create_cmap_asset(dirt, filepath, 'output-cmap-dirt', cmap_name='DIRT')
theia.create_cmap_asset(evi, filepath, 'output-cmap-evi')
## Dump to csv; pls remove if not needed. The size is huge
np.savetxt(filepath + "ndvi_dump.csv", ndvi, delimiter=",")
np.savetxt(filepath + "ndwi_dump.csv", ndwi, delimiter=",")
np.savetxt(filepath + "bai_dump.csv", bai, delimiter=",")
np.savetxt(filepath + "rvi_dump.csv", rvi, delimiter=",")
np.savetxt(filepath + "gndvi_dump.csv", gndvi, delimiter=",")
np.savetxt(filepath + "msavi_dump.csv", msavi, delimiter=",")
np.savetxt(filepath + "dirt_dump.csv", dirt, delimiter=",")
np.savetxt(filepath + "evi_dump.csv", evi, delimiter=",")
##
storage_url = gaia.store_asset_to_s3(filepath)
asset = Asset.objects.create_asset(aoi, {
'type': 'ANALYTIC',
'date': date,
'storage_url': storage_url,
'planet_item_id': planet_item_id,
'usability_score': usability_score,
'note': json.dumps({
'mean_ndvi': mean_ndvi,
'mean_ndwi': mean_ndwi,
'mean_bai': mean_bai,
'mean_rvi': mean_rvi,
'mean_gndvi': mean_gndvi,
'mean_msavi': mean_msavi,
'haze': is_hazy
})
})
return asset.id
