import gradio as gr
from src.pipeline.prediction_pipline import CustomData, PredictPipeline

# Constants
CUT_CHOICES = ["Fair", "Good", "Very Good", "Premium", "Ideal"]
COLOR_CHOICES = ["D", "E", "F", "G", "H", "I", "J"]
CLARITY_CHOICES = ["I1", "SI2", "SI1", "VS2", "VS1", "VVS2", "VVS1", "IF"]


def predict(inp_carat, inp_depth, inp_table, inp_x, inp_y, inp_z, selected_cut, selected_color, selected_clarity):
    """
    The function `predict` takes in various input parameters related to a diamond and uses a custom data
    object and a prediction pipeline to predict the price of the diamond.

    returns: A predicted diamond price rounded up to 2 decimal points.
    """
    data = CustomData(
        carat=inp_carat,
        depth=inp_depth,
        table=inp_table,
        x=inp_x,
        y=inp_y,
        z=inp_z,
        cut=selected_cut,
        color=selected_color,
        clarity=selected_clarity
    )

    parsed_data = data.get_data_as_dataframe()
    prediction_pipline = PredictPipeline()
    prediction = prediction_pipline.predict(parsed_data)

    return round(prediction[0], 2)


# A Gradio interface for the diamond price prediction.Defines the layout and components of the interface,
# such as input fields for carat, depth, table, x, y, z, cut, color, and clarity,
# A button for prediction and a textbox to display the predicted price.
# The `demo.launch()` statement launches the Gradio interface.
with gr.Blocks() as demo:
    gr.Markdown("# Welcome to Diamond Price Prediction.")
    gr.Markdown(
        "### For predicting the value enter the inputs and then click on **Predict** to see the result.")

    # with gr.Row():

    with gr.Row():
        carat = gr.Number(label="Carat")
        depth = gr.Number(label="Depth")
        table = gr.Number(label="Table")

    with gr.Row():
        x = gr.Number(label="X")
        y = gr.Number(label="Y")
        z = gr.Number(label="Z")

    with gr.Row():
        cut = gr.Dropdown(label="Cut", choices=CUT_CHOICES,
                          info="Diamond cut specifically refers to the quality of a diamond's angles, proportions, symmetrical facets, brilliance, fire, scintillation and finishing details.")
        color = gr.Dropdown(label="Color", choices=COLOR_CHOICES,
                            info="Diamond color is graded in terms of how white or colorless a diamond is.")
        clarity = gr.Dropdown(label="Clarity", choices=CLARITY_CHOICES,
                              info="A diamond's clarity grade evaluates how clean a diamond is from both inclusions and blemishes.")

    gr.Markdown("### Prediction: ")
    result = gr.Textbox(label="Predicted Price")
    predict_btn = gr.Button("Predict")
    predict_btn.click(fn=predict, inputs=[
                      carat, depth, table, x, y, z, cut, color, clarity], outputs=result)
    
    gr.Markdown("### Sample Example: Click on one the row to select the values.")
    examples = gr.Examples(examples = [
        [0.71, 61.4, 56, 5.74, 5.77, 3.53, "Ideal", "D", "VS2"],
        [2, 59.5, 57, 8.08, 8.15, 4.89, "Very Good", "G", "SI2"],
        [1.52, 60.8, 59, 7.36, 7.4, 4.49, "Premium", "G", "SI2"]
    ],inputs=[carat, depth, table, x, y, z, cut, color, clarity] )
    


demo.launch()