Developer Docs: What happens if you create a new project?

Creating a new project

After cloud_classifier class is initialized - see Developer Docs: What happens during Setup? - a new project can be created via

cc.create_new_project(name="NewRandomForestClassifier", path="../classifiers")

The arguments have the following meaning:

help( cc.create_new_project )

Help on method create_new_project in module cloud_project:

create_new_project(name, path=None) method of cloud_classifier.cloud_classifier instance
    Creates a persistant classifier project.
    
    
    Parameters
    ----------
    name : string
        Name of the the project that will be created
    
    path : string, optional
        Path to the directory where the project will be stored. If none is
        given, the current working directory will be used.

Three steps happen:

1. a new folder with name name is created under path

2. the parameter_handle class is used to initialize the new project settings

3. the new project is finally loaded

Initial new project settings via the parameter_handler class

This method is using a template mechanism. The template needs to be provided from somewhere.

In the standard setting, the template should be stored under the subfolder <package_path>/defaults. The variable parameter_handler.__default_path holds the information on the template path. The currently used content is:

> tree defaults/
defaults/
├── filelists
│   ├── evaluation_sets.json
│   ├── input_files.json
│   ├── label_files.json
│   └── training_sets.json
└── settings
    ├── config.json
    └── data_structure.json

2 directories, 6 files

The default setting files are copied over to the new project folder.