Python and netCDF sciprt do not operate anymore - variables

Question

I am using a python 2.6 script, that I have been using for quite a while now and I get an error that it shouldn't be there. The python script is run form the location of where the netCDF file is located, here is the code
from numpy import *
import numpy as numpy
from netCDF4 import Dataset
import datetime as DT
from time import strftime
import os
floc ='/media/USB-HDD/NCEP_NCAP data/data_2010/' #location of directory that the file resides
fname ='cfsr_Scotland_2010' # name of the netCDF file
in_ext = '.nc' # ending extentsion of the netCDF
basetime = DT.datetime(2010,01,01,0,0,0) # Initial time (start) for the netCDF
ncfile = Dataset(floc+fname+in_ext,'r') # netCDF assigned name
time = ncfile.variables['time']
lon = ncfile.variables['lon']
lat = ncfile.variables['lat']
uwind = ncfile.variables['10u']
vwind = ncfile.variables['10v']
ht = ncfile.variables['height']
I get the error in the ncfile naming, which is odd cause I checked the way its written
Traceback (most recent call last):
File "CFSR2WIND.py", line 24, in <module>
ncfile = Dataset(floc+fname+in_ext,'r') # netCDF assigned name
File "netCDF4.pyx", line 1317, in netCDF4.Dataset.__init__ (netCDF4.c:14608)
RuntimeError: No such file or directory
Does anybody know why and what caused this, and how can It be solved
thank you
george

Answer 1

Try using the netcdf module from scipy instead:
from scipy.io.netcdf import netcdf_file as Dataset
Couple other suggestions:
Importing numpy. You're importing it twice, and it's a bit dangerous to read in all instances using *. By convention, most people abbreviate numpy as np and load it as import numpy as np. Then you can call instances from numpy using np.mean() for example.
Concatenating the path, filename, and file extension. It's OK to use string concatenation using the + sign, but there is another way to do this using the join command. So, the total filename would be something like filename = ''.join([floc, fname, in_ext]).