Python Examples

All operations which are supported by the command line tool can also be performed in a more modular sense from python. In fact, actions can be defined more broadly when using the python interface. The use of the commandline interface is therefore not strictly necessary. The following sections show python examples which illustrate this. The Jupyter notebooks, which are used to build these examples can be downloaded from the github repository.

• Establish a database connection and retrieve the inventory and User settings
  • Read-only user versus priviliged user
  • Using geoslurpConnect
  • Alternatively one can use the class GeoslurpConnector and provide all connections details manually
  • Loading the Dataset Inventory
  • Setting and getting usersettings from the database
    • Default settings
    • Storing and retrieving Authentication details
• Downloading and registering “catalogue” datasets
• Download and register a dynamic dataset of (crowd-sourced?) US Marijuana street prices
  • Example: US street prices of Marijuana
    • Create the stucture of the database table using sqalchemy ORM methods
    • Create a helper function to extract relevant information from the file
    • Inherit from DataSet
    • Create an instance of the class and call the pull, and register methods to download and register the data
  • Avoid reinventing the wheel
• Discovery techniques
  • Query a single table and visualize the result
    • Pick a state and plot the results using matplotlib
  • A more complicated query which joins data between tables
    • Visualize the result geographically using matplotlib
• Exporting parts of the database to portable formats
  • Exporting a subset of the PSMSL table
  • Exporting an Argo query and the profile files in the South China Sea for the first 2 weeks of December 2010
• Exporting a selection of GRACE datasets and auxiliary datasets (without geographical information) to a local sqlite database.
• Applying the Ocean function to geometries and generating grids
  • Example 1: Generate a global 0.125 degree equidistant raster
  • Example 2: Use the optimized table for other queries
• Conclusion: An indexed and subdivided geometry table speeds up the ocean queries considerably