treesplantingsitesdataset.py

# -*- coding: utf-8 -*-
"""TreesPlantingSitesDataset

Automatically generated by Colaboratory.

Original file is located at
    https://colab.research.google.com/drive/1Hvt3Y131OjTl7oGQGS55S_v7-aYu1Yj8
"""

from datasets import DatasetBuilder, DownloadManager, DatasetInfo, SplitGenerator, Split
from datasets.features import Features, Value, ClassLabel
import pandas as pd
import geopandas as gpd
import matplotlib.pyplot as plt

class TreesPlantingSitesDataset(DatasetBuilder):
    VERSION = "1.0.0"

    def _info(self):
        # Specifies the dataset's features
        return DatasetInfo(
            description="This dataset contains information about tree planting sites from CSV and GeoJSON files.",
            features=Features({
                "OBJECTID": Value("int32"),  # Unique identifier for each record
                "streetaddress": Value("string"),  # Street address of the tree planting site
                "city": Value("string"),  # City where the tree planting site is located
                "zipcode": Value("int32"),  # Zip code of the tree planting site
                "facilityid": Value("int32"),  # Identifier for the facility
                "neighborhood": Value("string"),  # Neighborhood where the tree planting site is located
                "plantingwidth": Value("string"),  # Width available for planting
                "plantingcondition": Value("string"),  # Condition of the planting site
                "matureheight": Value("string"),  # Expected mature height of the tree
                "GlobalID": Value("string"),  # Global unique identifier
                "created_user": Value("string"),  # User who created the record
                "created_date": Value("string"),  # Date when the record was created
                "last_edited_user": Value("string"),  # User who last edited the record
                "last_edited_date": Value("string"),  # Date when the record was last edited
                "geometry": Value("string")  # Geometry feature from GeoJSON
            }),
            supervised_keys=None,
            homepage="https://github.com/AuraMa111?tab=repositories",
            citation="Citation for the dataset",
        )

    def _split_generators(self, dl_manager: DownloadManager):
        # Downloads the data and defines the splits
        urls_to_download = {
            "csv": "https://drive.google.com/uc?export=download&id=18HmgMbtbntWsvAySoZr4nV1KNu-i7GCy",
            "geojson": "https://drive.google.com/uc?export=download&id=1jpFVanNGy7L5tVO-Z_nltbBXKvrcAoDo"
        }
        downloaded_files = dl_manager.download_and_extract(urls_to_download)

        return [
            SplitGenerator(name=Split.TRAIN, gen_kwargs={
                "csv_path": downloaded_files["csv"],
                "geojson_path": downloaded_files["geojson"]
            }),
            # If you have additional splits, define them here
        ]

    def _generate_examples(self, csv_path, geojson_path):
        # Load the data into DataFrame and GeoDataFrame
        csv_data = pd.read_csv(csv_path)
        geojson_data = gpd.read_file(geojson_path)

        # Merge the CSV data with the GeoJSON data on the 'OBJECTID' column
        gdf = geojson_data.merge(csv_data, on='OBJECTID')
        columns_to_extract = [
            "OBJECTID", "streetaddress", "city", "zipcode", "facilityid", "present",
            "neighborhood", "plantingwidth", "plantingcondition", "underpowerlines",
            "matureheight", "GlobalID", "created_user", "created_date",
            "last_edited_user", "last_edited_date", "geometry"
        ]

        # Extract the specified columns
        extracted_gdf = gdf[columns_to_extract]
        # Basic statistics: Count the number of planting sites
        number_of_planting_sites = gdf['present'].value_counts()
        print("Number of planting sites:", number_of_planting_sites)

        # Spatial analysis: Group by neighborhood to see the distribution of features
        neighborhood_analysis = gdf.groupby('neighborhood').size()
        print("Distribution by neighborhood:", neighborhood_analysis)

        # Visual analysis: Plot the points on a map
        gdf.plot(marker='*', color='green', markersize=5)
        plt.title('TreesPlantingSitesDataset')
        plt.show()  # Make sure to display the plot if running in a script

        # Iterate over the rows in the GeoDataFrame and yield examples
        for id_, row in extracted_gdf.iterrows():
            yield id_, {
                "OBJECTID": row["OBJECTID"],
                "streetaddress": row["streetaddress"],
                "city": row["city"],
                "zipcode": row["zipcode"],
                "facilityid": row["facilityid"],
                "neighborhood": row["neighborhood"],
                "plantingwidth": row["plantingwidth"],
                "plantingcondition": row["plantingcondition"],
                "matureheight": row["matureheight"],
                "GlobalID": row["GlobalID"],
                "created_user": row["created_user"],
                "created_date": row["created_date"],
                "last_edited_user": row["last_edited_user"],
                "last_edited_date": row["last_edited_date"],
                "geometry": row["geometry"].wkt if row["geometry"] else None  # Ensure geometry is in Well-Known Text (WKT) format or handled as desired
            }