examples/notebooks/125_example_code.ipynb

{
    "cells": [
        {
            "cell_type": "markdown",
            "metadata": {},
            "source": [
                "<a href=\"https://githubtocolab.com/gee-community/geemap/blob/master/examples/notebooks/125_using_example_cde_.ipynb\" target=\"_parent\"><img src=\"https://colab.research.google.com/assets/colab-badge.svg\" alt=\"Open in Colab\"/></a>\n",
                "\n",
                "**Using Earth engine examples**\n",
                "\n",
                "\n",
                "- ee datasets seem to come from https://code.earthengine.google.com/scripts/load\n",
                "- When logged into gee, this link returns a f.json. It contains all code examples from the docs, including the datasets.\n",
                "- community datasets can be accessed using git from https://earthengine.googlesource.com\n",
                "- which in many cases can be translated into python code\n",
                "- this should make it feasible to use the code examples directly into geemap\n",
                "\n",
                "Uncomment the following line to install [geemap](https://geemap.org) if needed."
            ]
        },
        {
            "cell_type": "code",
            "execution_count": null,
            "metadata": {},
            "outputs": [],
            "source": [
                "# !pip install -U geemap"
            ]
        },
        {
            "cell_type": "code",
            "execution_count": null,
            "metadata": {},
            "outputs": [],
            "source": [
                "import ee\n",
                "import geemap\n",
                "import json"
            ]
        },
        {
            "cell_type": "markdown",
            "metadata": {},
            "source": [
                "## Manually download f.json \n",
                "https://code.earthengine.google.com/scripts/load"
            ]
        },
        {
            "cell_type": "markdown",
            "metadata": {},
            "source": [
                "## Explore json contents\n",
                "\n",
                "Change file download location. Also added f.json to repo."
            ]
        },
        {
            "cell_type": "code",
            "execution_count": null,
            "metadata": {},
            "outputs": [],
            "source": [
                "url = 'https://github.com/gee-community/geemap/blob/master/geemap/data/gee_f.json'\n",
                "geemap.download_file(url)"
            ]
        },
        {
            "cell_type": "code",
            "execution_count": null,
            "metadata": {},
            "outputs": [],
            "source": [
                "with open('gee_f.json', encoding=\"utf-8\") as f:\n",
                "    functions = json.load(f)\n",
                "\n",
                "functions.keys()"
            ]
        },
        {
            "cell_type": "code",
            "execution_count": null,
            "metadata": {},
            "outputs": [],
            "source": [
                "[(x['name'], len(x['contents'])) for x in functions['examples']]"
            ]
        },
        {
            "cell_type": "markdown",
            "metadata": {},
            "source": [
                "## Details"
            ]
        },
        {
            "cell_type": "code",
            "execution_count": null,
            "metadata": {},
            "outputs": [],
            "source": [
                "details = [\n",
                "    (dataset['name'], dataset['code'])\n",
                "    for x in functions['examples']\n",
                "    for dataset in x['contents']\n",
                "    if x['name'] == 'Datasets'\n",
                "]\n",
                "details[0]"
            ]
        },
        {
            "cell_type": "markdown",
            "metadata": {},
            "source": [
                "## Convert js to py"
            ]
        },
        {
            "cell_type": "code",
            "execution_count": null,
            "metadata": {},
            "outputs": [],
            "source": [
                "def get_py(js):\n",
                "    try:\n",
                "        return geemap.js_snippet_to_py(\n",
                "            js, add_new_cell=False, import_ee=False, import_geemap=False, show_map=False\n",
                "        )\n",
                "    except:\n",
                "        return None\n",
                "\n",
                "\n",
                "results_convert = [get_py(js) for _, js in details]"
            ]
        },
        {
            "cell_type": "code",
            "execution_count": null,
            "metadata": {},
            "outputs": [],
            "source": [
                "len(details), len(results_convert)"
            ]
        },
        {
            "cell_type": "markdown",
            "metadata": {},
            "source": [
                "Except for 2 cases, converts ran without error."
            ]
        },
        {
            "cell_type": "code",
            "execution_count": null,
            "metadata": {},
            "outputs": [],
            "source": [
                "len([x for x in results_convert if x])"
            ]
        },
        {
            "cell_type": "markdown",
            "metadata": {},
            "source": [
                "## Verify availability"
            ]
        },
        {
            "cell_type": "code",
            "execution_count": null,
            "metadata": {},
            "outputs": [],
            "source": [
                "from geemap import datasets\n",
                "\n",
                "len(datasets.get_ee_stac_list())"
            ]
        },
        {
            "cell_type": "code",
            "execution_count": null,
            "metadata": {},
            "outputs": [],
            "source": [
                "known_datasets = set(name.replace('/', '_') for name in datasets.get_ee_stac_list())\n",
                "known_code = set(name for name, _ in details)"
            ]
        },
        {
            "cell_type": "markdown",
            "metadata": {},
            "source": [
                "## Only some have exact matches"
            ]
        },
        {
            "cell_type": "code",
            "execution_count": null,
            "metadata": {},
            "outputs": [],
            "source": [
                "len(known_code & known_datasets)"
            ]
        },
        {
            "cell_type": "markdown",
            "metadata": {},
            "source": [
                "Sometimes there is a 2 example available for a dataset, mostly FeatureViews."
            ]
        },
        {
            "cell_type": "code",
            "execution_count": null,
            "metadata": {},
            "outputs": [],
            "source": [
                "excess = known_code - known_datasets"
            ]
        },
        {
            "cell_type": "code",
            "execution_count": null,
            "metadata": {},
            "outputs": [],
            "source": [
                "len(excess)"
            ]
        },
        {
            "cell_type": "code",
            "execution_count": null,
            "metadata": {},
            "outputs": [],
            "source": [
                "excess_no_featureview = set(x.replace('_FeatureView', '') for x in known_code)\n",
                "len(excess_no_featureview - known_datasets)"
            ]
        },
        {
            "cell_type": "markdown",
            "metadata": {},
            "source": [
                "Leftovers"
            ]
        },
        {
            "cell_type": "code",
            "execution_count": null,
            "metadata": {},
            "outputs": [],
            "source": [
                "(excess_no_featureview - known_datasets)"
            ]
        },
        {
            "cell_type": "markdown",
            "metadata": {},
            "source": [
                "## Datasets without examples"
            ]
        },
        {
            "cell_type": "code",
            "execution_count": null,
            "metadata": {},
            "outputs": [],
            "source": [
                "missing_code = known_datasets - known_code\n",
                "len(missing_code)"
            ]
        },
        {
            "cell_type": "markdown",
            "metadata": {},
            "source": [
                "These are also missing in gee itself (logically as these load from f.json). \n",
                "\n",
                "- https://code.earthengine.google.com/?scriptPath=Examples%3ADatasets%2FLANDSAT_LC8\n",
                "\n",
                "Deprecated stuff afaik, s.a. Landsat LT1"
            ]
        },
        {
            "cell_type": "code",
            "execution_count": null,
            "metadata": {},
            "outputs": [],
            "source": [
                "missing_code"
            ]
        },
        {
            "cell_type": "code",
            "execution_count": null,
            "metadata": {},
            "outputs": [],
            "source": [
                "from geemap import common\n",
                "\n",
                "[x['title'] for x in common.search_ee_data('LANDSAT L1T')][:5]"
            ]
        },
        {
            "cell_type": "code",
            "execution_count": null,
            "metadata": {},
            "outputs": [],
            "source": [
                "[(x['uid'], x['title']) for x in common.search_ee_data('MODIS MYD')][:5]"
            ]
        }
    ],
    "metadata": {
        "kernelspec": {
            "display_name": "Python 3",
            "language": "python",
            "name": "python3"
        }
    },
    "nbformat": 4,
    "nbformat_minor": 5
}