File size: 5,442 Bytes
a66cfba 2b0a4af a66cfba 2b0a4af a66cfba |
1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60 61 62 63 64 65 66 67 68 69 70 71 72 73 74 75 76 77 78 79 80 81 82 83 84 85 86 87 88 89 90 91 92 93 94 95 96 97 98 99 100 101 102 103 104 105 106 107 108 109 110 111 112 113 114 |
from typing import List, Dict
import httpx
import gradio as gr
import pandas as pd
import json
async def get_splits(dataset_name: str) -> Dict[str, List[Dict]]:
URL = f"https://huggingface.co/api/datasets/{dataset_name}"
async with httpx.AsyncClient() as session:
response = await session.get(URL)
dataset_info = response.json()
return {
"splits": [
{"split": split_name, "config": config_name}
for config_name, config_info in dataset_info.get("config", {}).items()
for split_name in config_info.get("splits", [])
]
}
async def get_first_rows(dataset: str, config: str, split: str) -> Dict[str, Dict[str, List[Dict]]]:
URL = f"https://huggingface.co/datasets/{dataset}/resolve/main/dataset_info.json"
async with httpx.AsyncClient() as session:
response = await session.get(URL)
dataset_info = response.json()
split_info = dataset_info["splits"][split]
first_rows = {
"rows": [
{"row": row} for row in split_info["examples"][:10]
]
}
return first_rows
# Guido von Roissum: https://www.youtube.com/watch?v=-DVyjdw4t9I
async def update_URL(dataset: str, config: str, split: str) -> str:
URL = f"https://huggingface.co/datasets/{dataset}/tree/main/{config}/{split}"
return URL
async def get_valid_datasets() -> List[str]:
URL = f"https://huggingface.co/api/datasets"
async with httpx.AsyncClient() as session:
response = await session.get(URL)
try:
datasets = [dataset["id"] for dataset in response.json()]
except (KeyError, json.JSONDecodeError):
datasets = [] # Set a default value if the response is not in the expected format
return datasets
def get_df_from_rows(api_output):
dfFromSort = pd.DataFrame([row["row"] for row in api_output["rows"]])
try:
dfFromSort.sort_values(by=1, axis=1, ascending=True, inplace=False, kind='mergesort', na_position='last', ignore_index=False, key=None)
except:
print("Exception sorting due to keyerror?")
return dfFromSort
async def update_configs(dataset_name: str):
splits = await get_splits(dataset_name)
all_configs = sorted(set([s["config"] for s in splits["splits"]]))
return (gr.Dropdown.update(choices=all_configs, value=all_configs[0]),
splits)
async def update_splits(config_name: str, state: gr.State):
splits_for_config = sorted(set([s["split"] for s in state["splits"] if s["config"] == config_name]))
dataset_name = state["splits"][0]["dataset"]
dataset = await update_dataset(splits_for_config[0], config_name, dataset_name)
return (gr.Dropdown.update(choices=splits_for_config, value=splits_for_config[0]), dataset)
async def update_dataset(split_name: str, config_name: str, dataset_name: str):
rows = await get_first_rows(dataset_name, config_name, split_name)
df = get_df_from_rows(rows)
return df
async def openurl(URL: str) -> str:
html = f"<a href={URL} target=_blank>{URL}</a>"
return (html)
with gr.Blocks() as demo:
gr.Markdown("<h1><center>🥫Datasetter📊 Datasets Analyzer and Transformer</center></h1>")
gr.Markdown("""<div align="center">Curated Datasets: <a href = "https://www.kaggle.com/datasets">Kaggle</a>. <a href="https://www.nlm.nih.gov/research/umls/index.html">NLM UMLS</a>. <a href="https://loinc.org/downloads/">LOINC</a>. <a href="https://www.cms.gov/medicare/icd-10/2022-icd-10-cm">ICD10 Diagnosis</a>. <a href="https://icd.who.int/dev11/downloads">ICD11</a>. <a href="https://paperswithcode.com/datasets?q=medical&v=lst&o=newest">Papers,Code,Datasets for SOTA in Medicine</a>. <a href="https://paperswithcode.com/datasets?q=mental&v=lst&o=newest">Mental</a>. <a href="https://paperswithcode.com/datasets?q=behavior&v=lst&o=newest">Behavior</a>. <a href="https://www.cms.gov/medicare-coverage-database/downloads/downloads.aspx">CMS Downloads</a>. <a href="https://www.cms.gov/medicare/fraud-and-abuse/physicianselfreferral/list_of_codes">CMS CPT and HCPCS Procedures and Services</a> """)
splits_data = gr.State()
with gr.Row():
dataset_name = gr.Dropdown(label="Dataset", interactive=True)
config = gr.Dropdown(label="Subset", interactive=True)
split = gr.Dropdown(label="Split", interactive=True)
with gr.Row():
#filterleft = gr.Textbox(label="First Column Filter",placeholder="Filter Column 1")
URLcenter = gr.Textbox(label="Dataset URL", placeholder="URL")
btn = gr.Button("Use Dataset")
#URLoutput = gr.Textbox(label="Output",placeholder="URL Output")
#URLoutput = gr.HTML(label="Output",placeholder="URL Output")
URLoutput = gr.HTML(label="Output")
with gr.Row():
dataset = gr.DataFrame(wrap=True, interactive=True)
demo.load(get_valid_datasets, inputs=None, outputs=[dataset_name])
dataset_name.change(update_configs, inputs=[dataset_name], outputs=[config, splits_data])
config.change(update_splits, inputs=[config, splits_data], outputs=[split, dataset])
split.change(update_dataset, inputs=[split, config, dataset_name], outputs=[dataset])
dataset_name.change(update_URL, inputs=[split, config, dataset_name], outputs=[URLcenter])
btn.click(openurl, [URLcenter], URLoutput)
demo.launch(debug=True) |