swantools
This repository gathers some useful functions and tools to be used alongside SWAN wave model.
For downloading, installing and using SWAN, please see: http://swanmodel.sourceforge.net/
To install the module, fork or download the source code and run:
sudo setup.py installThere are several dependencies that have to be met in order to use all funcions. The major ones are
- Numpy
- Pandas
- NetCDF4
- Matplotlib
- Scipy.io
Nowadays, there is a I/O class which can handle TABLE, SPECOUT and BLOCK outputs and write SPECOUT and TPAR files.
For example, to use the I/O structure for reading a TABLE made HEAD option do:
import swantools.io
R = swantools.io.SwanIO()
df = R.read_swantable('file.txt')This will return a pandas DataFrame with the model data.
For quickly plotting the results, just do:
import swantools.plot
import seaborn as sns
P = swantools.plot.SwanPlot()
y = df["Hsig"]
x = df.index.values
P.timeseries(x,y,"Significant Wave Heights")Reading spectral files (.spc) wrote with "SPECOUT SPEC2D" become very easy:
import swantools.io
R = swantools.io.SwanIO()
lon,lat,freqs,dirs,times,factors,spec = R.read_swanspc('file.spc')Plotting this data is also easy:
import swantools.plot
P = swantools.plot.SwanPlot()
P.spcplot(freqs,dirs,spectra)Block (.mat) files can be handled in both stationay and non-stationary versions. For extracting variables from this files, one can do:
import swantools.io
R = swantools.io.SwanIO()
lon,lat,times,hs = R.read_swanblock('block.mat','Hsig')Plotting field data can be done whith swantools.plot or creating a netCDF file and using ncview:
import swantools.plot
P = swantools.plot.SwanPlot()
P.blockplot(lon,lat,hs[0,:,:],"Non-stationary Results")or,
W = swantools.io.Converters()
W.np2nc("Hsig.nc",lat,lon,times,hs,"Significant Wave Height")
subprocess.call(ncview Hsig.nc,shell=True)Non-stationary two-dimensional spectral files (SPEC2D) can be saved as netCDF:
R = swantools.io.SwanIO()
W = swantools.io.Converters()
lon,lat,freqs,dirs,times,factors,spectrum = R.read_swanspc('spectrum.spc')
W.spc2nc("spectrum.nc",lat,lon,freqs,dirs,times,factors,spectrum)Non-stationary two-dimensional spectral files (SPEC2D) can be written as well:
# Getting some data to play with
R = swantools.io.SwanIO()
lat,lon,freqs,dirs,times,factors,spectrum = R.read_swanspc('spectrum.spc')
# Re-writing the data
R.write_spectrum("spcout.spc",lat,lon,times,freqs,dirs,factors,spectrum)Much more will be added in the future.
Enjoy !
