kalophain / lidaco

Python converter for Wind datasets.

Lidaco

Lidaco (Wind Lidar Data Converter) is a library and executable that enables a modular writing of data converters.

Following the configurations that are specified by the user on a config.yml(s), a Reader module is selected to import the data from input files. Similarly, a Writer is also selected to write the output file(s).

Lidaco works on datasets that can be described using the unidata Common Data Model. It can be used to process single files or entire folders.

Available Readers

* AQ500
* Galion
* WLS70
* Windcubev1
* Windcubev2
* Windscanner
* ZephIR300
* Triton

Available Writers

* MetadataCard
* NcML
* NetCDF4

Getting started

Install

pip install lidaco

Run in the command line (shell)

lidaco --config-file=samples/Windscanner/config.yaml

Use as a library

from lidaco.core.Builder import Builder

builder = Builder(config_file = 'path/to/config.yaml')
builder.build()

Converting data

Writing conversion files

Each conversion requires a configuration file written in YAML. This file contains four (optional) main groups of configurations:

Parameters

This section allows you to specify the parameters for the converter itself, the selected reader and or writer. The converter parameters are:

parameters:
  input: 
    path: ./path/to/input/folder/
    format: Windscanner                 # The name should match 
  output: 
    path: ./path/to/output/folder       # optional, defaults to ./output/
    format: NetCDF4
  
  # Optional, and specifies the number of input files to be concatenated per output file.
  # If not specified defaults to 1 (1 output per input file). The last output will contain <= output_block_size input files.
  # If the parameter is specified with None (the same as without any value), all input files will be concatenated into a single output file.
  output_block_size: 3

Some of these parameters can be overridden using the command arguments, run lidaco --help to know them.

When specifying relative paths, the configuration file location wherein these are specified, or the current directory when specifying them as command arguments, is the one used to resolve its absolute value.

Additional parameters might be necessary depending on the specific reader/writer in use. Read the respective module documentation available, or try to take a look at its source code.

Attributes

This section specify the global attributes that will be added to the dataset. You can add all attributes you desire.

attributes:
  # e.g.,
  lidar_technology: 'pulsed'
  lidar_scanning_type: 'vertical profiling'
  data_processing_history: 'data taken from .scn files generated by ...'

Variables

This section is similar to Attributes but instead of reading being used to specify Attributes, it is used to specify variables.

variables:   
  # e.g.,
  pitch:
    data_type: 'f4'
    units: 'degrees'
    long_name: 'lidar_pitch_angle'
    comment: ''
    accuracy: ''
    accuracy_info: 'No information on pitch accuracy available.'
    value: 0

Imports

Lidaco configuration system is built to motivate the configurations reusability, so each config file can import others. With this mechanism you can to split your configurations into devices, scenarios, campaigns and so on. In case of multiple definitions of an attribute or variable, the definitions in the file prevail over the others that are imported.

imports: # read in order
  - ./general/NEWA_Kassel_general_dataset.yaml
  - ./instruments/Windcubev2_general_instrument_description.yaml
  - ./processing/NEWA_Kassel_data_processing_history.yaml

For more examples on how to setup your config files, take a look at the available Samples.

---

Contributing

If you would like to add, or see being added, some change to the converter you can:

• Open an issue so that it is discussed by everyone.
• And/or also Submit a pull request with the desired changes. (About pull requests)

To submit a new Reader/Writer

• Create a python class at lidaco.readers or lidaco.writers, if you are writing a reader or writer respectively.
• The file and class should have the same name, that will be used in the config files.
• It should extend core.reader or core.writer.

Take a look at the existing readers and writers.