whackthejacker/DataHubHub

1import streamlit as st2import pandas as pd3import json4import io5from utils.dataset_utils import get_dataset_info, detect_dataset_format6 7def render_dataset_uploader():8    """9    Renders the dataset upload component that supports CSV and JSON formats.10    """11    st.markdown("""12    <div class="upload-container">13        <p>Upload your dataset in CSV or JSON format</p>14    </div>15    """, unsafe_allow_html=True)16    17    # File uploader18    uploaded_file = st.file_uploader(19        "Choose a file", 20        type=["csv", "json"], 21        help="Upload a CSV or JSON file containing your dataset"22    )23    24    # Sample dataset option25    st.markdown("Or use a sample dataset:")26    sample_dataset = st.selectbox(27        "Select a sample dataset",28        ["None", "Iris Dataset", "Titanic Dataset", "Boston Housing Dataset"]29    )30    31    # Process uploaded file32    if uploaded_file is not None:33        try:34            # Check file extension35            file_extension = uploaded_file.name.split(".")[-1].lower()36            37            if file_extension == "csv":38                df = pd.read_csv(uploaded_file)39                dataset_type = "csv"40            elif file_extension == "json":41                # Try different JSON formats42                try:43                    # First try parsing as a regular JSON with records orientation44                    df = pd.read_json(uploaded_file)45                    dataset_type = "json"46                except:47                    # If that fails, try to parse as JSON Lines48                    try:49                        df = pd.read_json(uploaded_file, lines=True)50                        dataset_type = "jsonl"51                    except:52                        # If that also fails, load raw JSON and convert53                        content = json.loads(uploaded_file.getvalue().decode("utf-8"))54                        if isinstance(content, list):55                            df = pd.DataFrame(content)56                        elif isinstance(content, dict):57                            # Handle nested dict structures58                            if any(isinstance(v, list) for v in content.values()):59                                # Find the list field and use it60                                for key, value in content.items():61                                    if isinstance(value, list):62                                        df = pd.DataFrame(value)63                                        break64                            else:65                                # Flat dict or dict of dicts66                                df = pd.DataFrame([content])67                        dataset_type = "json"68            else:69                st.error(f"Unsupported file format: {file_extension}")70                return71            72            # Store dataset and its info in session state73            st.session_state.dataset = df74            st.session_state.dataset_name = uploaded_file.name75            st.session_state.dataset_type = dataset_type76            st.session_state.dataset_info = get_dataset_info(df)77            78        except Exception as e:79            st.error(f"Error loading dataset: {str(e)}")80    81    # Process sample dataset82    elif sample_dataset != "None":83        try:84            if sample_dataset == "Iris Dataset":85                # Load Iris dataset86                from sklearn.datasets import load_iris87                iris = load_iris()88                df = pd.DataFrame(data=iris.data, columns=iris.feature_names)89                df['target'] = iris.target90                dataset_type = "csv"91                92            elif sample_dataset == "Titanic Dataset":93                # URL for Titanic dataset94                url = "https://raw.githubusercontent.com/datasciencedojo/datasets/master/titanic.csv"95                df = pd.read_csv(url)96                dataset_type = "csv"97                98            elif sample_dataset == "Boston Housing Dataset":99                # Load Boston Housing dataset100                from sklearn.datasets import fetch_california_housing101                housing = fetch_california_housing()102                df = pd.DataFrame(data=housing.data, columns=housing.feature_names)103                df['target'] = housing.target104                dataset_type = "csv"105            106            # Store dataset and its info in session state107            st.session_state.dataset = df108            st.session_state.dataset_name = sample_dataset109            st.session_state.dataset_type = dataset_type110            st.session_state.dataset_info = get_dataset_info(df)111            112        except Exception as e:113            st.error(f"Error loading sample dataset: {str(e)}")114