Add a function to simplify setting up Colab (install packages, authenticate).

When expanding a node for the first time, the JSON blob is briefly exposed before parsed to tree, and if long, extends the panel across the whole map.

The tree building and opening handler are a lot of code and maybe more complicated than needed. Is the lazy loading beneficial when there are 1000's of elements vs rendering the entire tree? A foundation for building the entire tree follows. It still needs to be tested against objects with 1000's of elements to see how fast it will render and respond. If it is performant, we should maybe consider using it, to simplify and avoid exposed JSON. If it is slow to render and not very responsive, we need to figure out how to not expose the JSON in the current implementation.

!pip install -q ipytree
from ipytree import Tree, Node
from google.colab import output
output.enable_custom_widget_manager()

def create_tree(data, name=None):
    if name is None:
        name = type(data).__name__
    node = Node(f'{name}: ({len(data)} elements)', show_icon=False, opened=False)
    if isinstance(data, dict):
        for key, value in data.items():
            if isinstance(value, (dict, list)):
                node.add_node(create_tree(value, key))
            else:
                child_node = Node(f'{key}: {value}', show_icon=False)
                node.add_node(child_node)
    elif isinstance(data, list):
        for index, item in enumerate(data):
            if isinstance(item, (dict, list)):
                node.add_node(create_tree(item, f'{index}'))
            else:
                child_node = Node(f'{index}: {item}', show_icon=False)
                node.add_node(child_node)
    return node

data = {
  "type": "Image",
  "bands": [
    {
      "id": "B1",
      "data_type": {
        "type": "PixelType",
        "precision": "float"
      },
      "dimensions": [
        9161,
        9161
      ],
      "crs": "EPSG:32628",
      "crs_transform": [
        30,
        0,
        341085,
        0,
        -30,
        8808015
      ]
    },
    {
      "id": "B2",
      "data_type": {
        "type": "PixelType",
        "precision": "float"
      },
      "dimensions": [
        9161,
        9161
      ],
      "crs": "EPSG:32628",
      "crs_transform": [
        30,
        0,
        341085,
        0,
        -30,
        8808015
      ]
    },
  ]
}

tree = Tree(animation=0)
tree.add_node(create_tree(data))
tree

A widget to monitor the status of ee.Tasks would be a good fit for this package IMO. An implementation using ipywidget can be found here: TaskManager from ipygee. Would be happy to contribute if I find some free time (a bit difficult at the moment, but could happen in a near future)

See:
https://ipywidgets.readthedocs.io/en/8.0.1/examples/Widget%20Asynchronous.html

earthengine_jupyter -> ee_jupyter

An initial issue intended to exercising the changelog, release, and push to PyPI.

For example, mimicking the Earth Engine Code Editor.

Similar to the Code Editor, maintain the state of the inspector tree when the updating information.

Allow initialization to override the default folders (Point, Pixels, Objects) used by the inspector.

eerepr

ipyleaflet.Map.center expects inputs to be ordered as (lat, lon). This is the opposite of the convention used by GeoJSON and Earth Engine.

Suggest: center_lon_lat()

Mimic the Earth Engine ui.Chart module but for Jupyter

Example:

When trying to sync two maps there is an error message that "method not found"

Even after importing all modules

Notebook to reproduce issue: https://colab.research.google.com/drive/1PLR3XzX9bk0N3_olRGOz1FfSAjvhgtKm#scrollTo=WQWutJtAtMm4

Screenshot:

This may be addressed by using the Asynchronous Widget generator approach:
https://ipywidgets.readthedocs.io/en/stable/examples/Widget%20Asynchronous.html#Generator-approach

Simiar to https://developers.google.com/earth-engine/apidocs/ui-map-centerobject