app.py

from datasets import load_dataset, IterableDataset
from functools import partial
from pandas import DataFrame
import gradio as gr 
import numpy as np
import tqdm
import json
import os

DEBUG = False

sets = {
    "satellogic": {
        "shards" : 3676,
    },
    "sentinel_1": {
        "shards" : 1763,
    },
    "neon": {
        "config" : "default",
        "shards" : 607,
        "path"   : "data",
    }
}

def open_dataset(dataset, set_name, split, batch_size, shard = -1):
    # I should really move ds.config_name and dsi to a gr.State()
    global dsi, ds

    if shard == -1:
        data_files = None
        shards = 100
    else:
        config = sets[set_name].get("config", set_name)
        shards = sets[set_name]["shards"]
        path   = sets[set_name].get("path", set_name)
        data_files = {"train":[f"{path}/{split}-{shard:05d}-of-{shards:05d}.parquet"]}

    if DEBUG:
        ds  = lambda:None
        ds.n_shards = 1234
        dsi = range(100)
    else:
        ds = load_dataset(
            dataset,
            config,
            split=split,
            cache_dir="dataset",
            data_files=data_files,
            streaming=True,
            token=os.environ.get("HF_TOKEN", None))
    
        dsi = iter(ds)

    return (
        gr.update(label=f"Shards (max {shards})", value=shard, maximum=shards),
        *get_images(batch_size)
    )
    
def get_images(batch_size):
    global dsi, ds
        
    items = []
    metadatas = []

    for i in tqdm.trange(batch_size, desc=f"Getting images"):
        if DEBUG:
            image = np.random.randint(0,255,(384,384,3))
            metadata = {"bounds":[[1,1,4,4]], }
        else:
            try:
                item = next(dsi)
            except StopIteration:
                break
            metadata = item["metadata"]
            if ds.config_name == "satellogic":
                # image = (np.asarray(item["1m"])).astype("uint8")
                # items.append(image[0,0,:,:])
                image = np.asarray(item["rgb"][0]).astype(np.uint8)
                items.append(image.transpose(1,2,0))

            if ds.config_name == "sentinel_1":
                metadata = json.loads(metadata)
                data = np.asarray(item["10m"])
                for i in range(data.shape[0]):
                    # Mapping of V and H to RGB. May not be correct
                    # https://gis.stackexchange.com/questions/400726/creating-composite-rgb-images-from-sentinel-1-channels
                    image = np.zeros((3,384,384), "uint8")
                    image[0] = data[i][0]
                    image[1] = data[i][1]
                    image[2] = (image[0]/(image[1]+0.1))*256
                    items.append(image.transpose(1,2,0))

            if ds.config_name == "default":
                dataRGB = np.asarray(item["rgb"]).astype("uint8")
                dataCHM = np.asarray(item["chm"]).astype("uint8")
                data1m  = np.asarray(item["1m"]).astype("uint8")
                for i in range(dataRGB.shape[0]):
                    image = dataRGB[i,:,:,:]
                    items.append(image.transpose(1,2,0))

                    image = dataCHM[i,0,:,:]
                    items.append(image)

                    image = data1m[i,0,:,:]
                    items.append(image)
            metadatas.append(metadata)

    return items, DataFrame(metadatas)

def skip(count, batch_size):
    global dsi

    skip = count*batch_size

    gr.Info(f"Skipping {skip} images (it's slow)")
    for i in tqdm.trange(skip, desc=f"Skipping {skip} images"):
        if DEBUG:
            pass
        else:
            next(dsi)
    
    return get_images(batch_size)

def update_shape(rows, columns):
    return gr.update(rows=rows, columns=columns)


with gr.Blocks(title="Dataset Explorer", fill_height = True) as demo:
    gr.Markdown("# [satellogic/EarthView](https://huggingface.co/datasets/satellogic/EarthView) Dataset Viewer")
    batch_size = gr.Number(10, label = "Batch Size", render=False)
    shard = gr.Slider(label="Shard", minimum=0, maximum=10000, step=1, render=False)
    table = gr.DataFrame(render = False)
    # headers=["Index","TimeStamp","Bounds","CRS"], 

    gallery = gr.Gallery(
        label="satellogic/EarthView",
        interactive=False,
        columns=5, rows=2, render=False)

    with gr.Row():
        dataset = gr.Textbox(label="Dataset", value="satellogic/EarthView")
        config = gr.Dropdown(choices=["satellogic", "sentinel_1", "neon"], label="Subset", value="satellogic", )
        split = gr.Textbox(label="Split", value="train")
        initial_shard = gr.Number(label = "Initial shard", value=0)

        gr.Button("Load (minutes)").click(
            open_dataset,
            inputs=[dataset, config, split, batch_size, initial_shard],
            outputs=[shard, gallery, table])

    gallery.render()
    
    with gr.Row():
        batch_size.render()

        rows = gr.Number(2, label="Rows")
        columns = gr.Number(5, label="Coluns")

        rows.change(update_shape, [rows, columns], [gallery])
        columns.change(update_shape, [rows, columns], [gallery])

    with gr.Row():
        shard.render()
        shard.release(
            open_dataset,
            inputs=[dataset, config, split, batch_size, shard],
            outputs=[shard, gallery, table])

        btn = gr.Button("Get More Images", scale=0)
        btn.click(get_images, [batch_size], [gallery, table])
        btn.click()

        # btn = gr.Button("Skip 10 Batches", scale=0)
        # btn.click(partial(skip, 10), [batch], gallery)

        # btn = gr.Button("Skip 25 Batches", scale=0)
        # btn.click(partial(skip, 25), [batch], gallery)
    
    table.render()

demo.launch(show_api=False)