Argo data python library

argopy is a python library that aims to ease Argo data access, visualisation and manipulation for regular users as well as Argo experts and operators.

Several python packages exist: we are currently in the process of analysing how to merge these libraries toward a single powerfull tool.
List your tool here !

Click here to and play with argopy before you even install it (thanks Pangeo).

Install

Install the last release with pip:

pip install argopy

But since this is a young library in active development, use direct install from this repo to benefit from the lastest version:

pip install git+http://github.com/euroargodev/argopy.git@master

The argopy library should work under all OS (Linux, Mac and Windows) and with python versions 3.6, 3.7 and 3.8.

Usage

Fetching Argo Data

Init the default data fetcher like:

from argopy import DataFetcher as ArgoDataFetcher
argo_loader = ArgoDataFetcher()

and then, request data for a specific space/time domain:

ds = argo_loader.region([-85,-45,10.,20.,0,10.]).to_xarray()
ds = argo_loader.region([-85,-45,10.,20.,0,1000.,'2012-01','2012-12']).to_xarray()

for profiles of a given float:

ds = argo_loader.profile(6902746, 34).to_xarray()
ds = argo_loader.profile(6902746, np.arange(12,45)).to_xarray()
ds = argo_loader.profile(6902746, [1,12]).to_xarray()

or for one or a collection of floats:

ds = argo_loader.float(6902746).to_xarray()
ds = argo_loader.float([6902746, 6902747, 6902757, 6902766]).to_xarray()

By default fetched data are returned in memory as xarray.DataSet. From there, it is easy to convert it to other formats like a Pandas dataframe:

ds = ArgoDataFetcher().profile(6902746, 34).to_xarray()
df = ds.to_dataframe()

or to export it to files:

ds = argo_loader.region([-85,-45,10.,20.,0,100.]).to_xarray()
ds.to_netcdf('my_selection.nc')
# or by profiles:
ds.argo.point2profile().to_netcdf('my_selection.nc')

Argo Index Fetcher

Index object is returned as a pandas dataframe.

Init the fetcher:

    from argopy import IndexFetcher as ArgoIndexFetcher

    index_loader = ArgoIndexFetcher()
    index_loader = ArgoIndexFetcher(backend='erddap')    
    #Local ftp backend 
    #index_loader = ArgoIndexFetcher(backend='localftp',path_ftp='/path/to/your/argo/ftp/',index_file='ar_index_global_prof.txt')

and then, set the index request index for a domain:

    idx=index_loader.region([-85,-45,10.,20.])
    idx=index_loader.region([-85,-45,10.,20.,'2012-01','2014-12'])

or for a collection of floats:

    idx=index_loader.float(6902746)
    idx=index_loader.float([6902746, 6902747, 6902757, 6902766])   

then you can see you index as a pandas dataframe or a xarray dataset :

    idx.to_dataframe()
    idx.to_xarray()

For plottings methods, you'll need matplotlib, cartopy and seaborn installed (they're not in requirements).
For plotting the map of your query :

    idx.plot('trajectory)    

For plotting the distribution of DAC or profiler type of the indexed profiles :

    idx.plot('dac')    
    idx.plot('profiler')`

Development roadmap

We aim to provide high level helper methods to load Argo data and meta-data from:

• Ifremer erddap
• local copy of the GDAC ftp folder
• Index files
• any other usefull access point to Argo data ?

We also aim to provide high level helper methods to visualise and plot Argo data and meta-data:

• Map with trajectories
• Waterfall plots
• T/S diagram
• etc !