def extract_data(delta=3):
"""
This is the master function that calls subsequent functions
to extract uwnd, vwnd, slp for the specified
tide gauges
delta: distance (in degrees) from the tide gauge
"""
print('Delta = {}'.format(delta), '\n')
#defining the folders for predictors
dir_in = "G:\\04_merra\\merraNetCDF"
surge_path = "D:\data\obs_surge"
csv_path = "G:\\04_merra\\merraNewLocalized"
#cd to the obs_surge dir to get TG information
os.chdir(surge_path)
tg_list = os.listdir()
#cd to the obs_surge dir to get TG information
os.chdir(dir_in)
years = os.listdir()
#################################
#looping through the year folders
#################################
for yr in years:
os.chdir(dir_in)
print(yr, '\n')
os.chdir(os.path.join(dir_in, yr))
####################################
#looping through the daily .nc files
####################################
for dd in os.listdir():
os.chdir(os.path.join(dir_in, yr)) #back to the predictor folder
print(dd, '\n')
#########################################
#get netcdf components - predictor file
#########################################
nc_file = readnetcdf(dd)
lon, lat, time, predSLP, predU10, predV10 = \
nc_file[0], nc_file[1], nc_file[2], nc_file[3], nc_file[4]\
, nc_file[5]
#looping through individual tide gauges
for t in range(0, len(tg_list)):
#the name of the tide gauge - for saving purposes
# tg = tg_list[t].split('.mat.mat.csv')[0]
tg = tg_list[t]
#extract lon and lat data from surge csv file
print(tg, '\n')
os.chdir(surge_path)
if os.stat(tg).st_size == 0:
print('\n', "This tide gauge has no surge data!", '\n')
continue
surge = pd.read_csv(tg, header=None)
#surge_with_date = add_date(surge)
#define tide gauge coordinate(lon, lat)
tg_cord = Coordinate(surge.iloc[0, 0], surge.iloc[0, 1])
#find closest grid points and their indices
close_grids = findPixels(tg_cord, delta, lon, lat)
ind_grids = findindx(close_grids, lon, lat)
#loop through preds#
#subset predictor on selected grid size
predictors = {'slp':predSLP, 'wnd_u':predU10, \
'wnd_v':predV10}
for xx in predictors.keys():
start_time = tt.time()
pred_new = subsetter(dd, predictors[xx], ind_grids, time)
print("--- %s seconds ---" % (tt.time() - start_time))
#create directories to save pred_new
os.chdir(csv_path)
#tide gauge directory
tg_name = tg.split('.mat.mat.csv')[0]
try:
os.makedirs(tg_name)
os.chdir(tg_name) #cd to it after creating it
except FileExistsError:
#directory already exists
os.chdir(tg_name)
#predictor directory
pred_name = xx
try:
os.makedirs(pred_name)
os.chdir(pred_name) #cd to it after creating it
except FileExistsError:
#directory already exists
os.chdir(pred_name)
#time for saving file
ncTime = dd.split('.')[2]
#save as csv
save_name = '_'.join([tg_name, pred_name, ncTime])\
+ ".csv"
pred_new.to_csv(save_name)
def extract_data(delta=3):
"""
This is the master function that calls subsequent functions
to extract uwnd, vwnd, slp for the specified
tide gauges
delta: distance (in degrees) from the tide gauge
"""
print('Delta = {}'.format(delta), '\n')
#defining the folders for predictors
dir_in = "/lustre/fs0/home/mtadesse/MERRAv2/data"
surge_path = "/lustre/fs0/home/mtadesse/obs_surge"
csv_path = "/lustre/fs0/home/mtadesse/merraLocalized"
#cd to the obs_surge dir to get TG information
os.chdir(surge_path)
tg_list = os.listdir()
#cd to the obs_surge dir to get TG information
os.chdir(dir_in)
years = os.listdir()
#################################
#looping through the year folders
#################################
#to mark the first csv
firstCsv = True
for yr in years:
os.chdir(dir_in)
#print(yr, '\n')
os.chdir(os.path.join(dir_in, yr))
####################################
#looping through the daily .nc files
####################################
for dd in os.listdir():
os.chdir(os.path.join(dir_in, yr)) #back to the predictor folder
print(dd, '\n')
#########################################
#get netcdf components - predictor file
#########################################
nc_file = readnetcdf(dd)
lon, lat, time, predSLP, predU10, predV10 = \
nc_file[0], nc_file[1], nc_file[2], nc_file[3], nc_file[4]\
, nc_file[5]
x = 376
y = 377
#looping through individual tide gauges
for t in range(x, y):
#the name of the tide gauge - for saving purposes
# tg = tg_list[t].split('.mat.mat.csv')[0]
tg = tg_list[t]
#extract lon and lat data from surge csv file
#print(tg, '\n')
os.chdir(surge_path)
if os.stat(tg).st_size == 0:
print('\n', "This tide gauge has no surge data!", '\n')
continue
surge = pd.read_csv(tg, header=None)
#surge_with_date = add_date(surge)
#define tide gauge coordinate(lon, lat)
tg_cord = Coordinate(surge.iloc[0, 0], surge.iloc[0, 1])
#find closest grid points and their indices
close_grids = findPixels(tg_cord, delta, lon, lat)
ind_grids = findindx(close_grids, lon, lat)
#loop through preds#
#subset predictor on selected grid size
predictors = {'slp':predSLP, 'wnd_u':predU10, \
'wnd_v':predV10}
for xx in predictors.keys():
pred_new = subsetter(dd, predictors[xx], ind_grids, time)
if xx == 'slp':
if firstCsv:
finalSLP = pred_new
else:
finalSLP = pd.concat([finalSLP, pred_new], axis=0)
print(finalSLP.shape)
elif xx == 'wnd_u':
if firstCsv:
finalUwnd = pred_new
else:
finalUwnd = pd.concat([finalUwnd, pred_new],
axis=0)
elif xx == 'wnd_v':
if firstCsv:
finalVwnd = pred_new
firstCsv = False
else:
finalVwnd = pd.concat([finalVwnd, pred_new],
axis=0)
#create directories to save pred_new
os.chdir(csv_path)
#tide gauge directory
tg_name_old = tg.split('.mat.mat.csv')[0]
tg_name = '-'.join([str(t), tg_name_old])
try:
os.makedirs(tg_name)
os.chdir(tg_name) #cd to it after creating it
except FileExistsError:
#directory already exists
os.chdir(tg_name)
#save as csv
finalSLP.to_csv('slp.csv')
finalUwnd.to_csv('wnd_u.csv')
finalVwnd.to_csv('wnd_v.csv')
def extract_data(delta=3):
"""
This is the master function that calls subsequent functions
to extract uwnd, vwnd, slp for the specified
tide gauges
delta: distance (in degrees) from the tide gauge
"""
print('Delta = {}'.format(delta), '\n')
#defining the folders for predictors
dir_in = "/lustre/fs0/home/mtadesse/MERRAv2/data"
surge_path = "/lustre/fs0/home/mtadesse/obs_surge"
csv_path = "/lustre/fs0/home/mtadesse/merraLocalized"
#cd to the obs_surge dir to get TG information
os.chdir(surge_path)
tg_list = os.listdir()
#cd to the obs_surge dir to get TG information
os.chdir(dir_in)
years = os.listdir()
#################################
#looping through the year folders
#################################
for yr in years:
os.chdir(dir_in)
print(yr, '\n')
os.chdir(os.path.join(dir_in, yr))
####################################
#looping through the daily .nc files
####################################
for dd in os.listdir():
os.chdir(os.path.join(dir_in, yr)) #back to the predictor folder
print(dd, '\n')
#########################################
#get netcdf components - predictor file
#########################################
nc_file = readnetcdf(dd)
lon, lat, time, predSLP, predU10, predV10 = \
nc_file[0], nc_file[1], nc_file[2], nc_file[3], nc_file[4]\
, nc_file[5]
x = startVal
y = endVal
#looping through individual tide gauges
for t in range(x, y):
#the name of the tide gauge - for saving purposes
# tg = tg_list[t].split('.mat.mat.csv')[0]
tg = tg_list[t]
#extract lon and lat data from surge csv file
print(tg, '\n')
os.chdir(surge_path)
if os.stat(tg).st_size == 0:
print('\n', "This tide gauge has no surge data!", '\n')
continue
surge = pd.read_csv(tg, header=None)
#surge_with_date = add_date(surge)
#define tide gauge coordinate(lon, lat)
tg_cord = Coordinate(surge.iloc[0, 0], surge.iloc[0, 1])
#find closest grid points and their indices
close_grids = findPixels(tg_cord, delta, lon, lat)
ind_grids = findindx(close_grids, lon, lat)
#loop through preds#
#subset predictor on selected grid size
predictors = {'slp':predSLP, 'wnd_u':predU10, \
'wnd_v':predV10}
for xx in predictors.keys():
pred_new = subsetter(dd, predictors[xx], ind_grids, time)
#create directories to save pred_new
os.chdir(csv_path)
#tide gauge directory
tg_name_old = tg.split('.mat.mat.csv')[0]
tg_name = '-'.join([str(t), tg_name_old])
try:
os.makedirs(tg_name)
os.chdir(tg_name) #cd to it after creating it
except FileExistsError:
#directory already exists
os.chdir(tg_name)
#predictor directory
pred_name = xx
try:
os.makedirs(pred_name)
os.chdir(pred_name) #cd to it after creating it
except FileExistsError:
#directory already exists
os.chdir(pred_name)
#time for saving file
ncTime = dd.split('.')[2]
#save as csv
save_name = '_'.join([tg_name, pred_name, ncTime])\
+ ".csv"
pred_new.to_csv(save_name)
#concatenate and delete csv on the fly
extension = 'csv'
all_filenames = [
i for i in glob.glob('*.{}'.format(extension))
]
#combine all files in the list
combined_csv = pd.concat(
[pd.read_csv(f) for f in all_filenames])
#export to csv
combined_csv.to_csv('.'.join([xx, 'csv']),
index=False,
encoding='utf-8-sig')
#delete the individual file
for file in os.listdir():
if file.startswith(str(t)):
os.remove(file)
