Delete Feb29 fields from a netcdf file

del_feb29.py

import os
from netCDF4 import Dataset, num2date
import calendar
import numpy as np

def del_feb29_in_file(in_path, varname):
	#input file
	dsin = Dataset(in_path)
 
	if not os.path.isdir(varname):
		os.mkdir(varname)

	out_path = "{}/{}".format(varname, os.path.basename(in_path)) 


	if os.path.isfile(out_path):
		dsin.close()
		print "{} already exists, will not redo!".format(out_path)
		return

	print "processing : {}".format(in_path)
	
	#output file
	dsout = Dataset(out_path, "w", format="NETCDF3_CLASSIC")
 
	#Copy dimensions
	for dname, the_dim in dsin.dimensions.iteritems():
    		#print dname, len(the_dim)
    		dsout.createDimension(dname, len(the_dim) if not the_dim.isunlimited() else None)


	# Delete all fileds on Feb 29
	time_var = dsin.variables["time"]
	print time_var[:].min(), time_var[:].max()
	dates = num2date(time_var[:], time_var.units)
	
	time_indices = np.where([not (d.month == 2 and d.day == 29) for d in dates])
 
 
	#Copy variables
	for v_name, varin in dsin.variables.iteritems():
    		out_var = dsout.createVariable(v_name, varin.datatype, varin.dimensions)
    		
		print v_name

		#print varin.datatype
		if "time" not in varin.dimensions:
	    		out_var[:] = varin[:]
		else:
	    		took = varin[:].take(time_indices, axis=varin.dimensions.index("time"))
			print took.shape
			out_var[:] = took.squeeze()
		out_var.setncatts({k: varin.getncattr(k) for k in varin.ncattrs()})
	
	#close the output file
	dsout.close()
	dsin.close()




def main():
	in_folder = "../ERA-Interim_0.75_NEMO_pilot"
	
	for vname in os.listdir(in_folder):
		for fname in os.listdir(os.path.join(in_folder, vname)):
			del_feb29_in_file(os.path.join(in_folder, vname, fname), vname)
			

 


if __name__ == "__main__":
	main()