import streamlit as st import pandas as pd from model import predict_kara_layout st.set_page_config(page_title="KARA II Layout – Regression v1", layout="wide") st.title("KARA II System Design – Regression Model v1") st.markdown( """ This app uses a **tiny regression model** trained on synthetic L'Acoustics KARA II layouts. - Input **room dimensions** and **boxes per side** - The model predicts **X, Y, Z, Ver, Hor, Rot** for each element of the L/R arrays - Download results as CSV for use in EASE / Soundvision / spreadsheets """ ) with st.sidebar: st.header("Room Parameters") room_length = st.number_input("Room Length (ft)", min_value=40.0, max_value=300.0, value=100.0, step=1.0) room_width = st.number_input("Room Width (ft)", min_value=30.0, max_value=200.0, value=80.0, step=1.0) room_height = st.number_input("Room Height (ft)", min_value=15.0, max_value=80.0, value=40.0, step=1.0) st.header("Array Parameters") boxes_per_side = st.number_input("Boxes per Side", min_value=4, max_value=15, value=15, step=1) system_id = st.number_input("System ID", min_value=1, max_value=9999, value=1, step=1) generate = st.button("Generate Layout (ML Model)") if generate: rows = predict_kara_layout( room_length=float(room_length), room_width=float(room_width), room_height=float(room_height), boxes_per_side=int(boxes_per_side), system_id=int(system_id), ) df_layout = pd.DataFrame(rows) st.subheader("Predicted Layout Table") st.dataframe(df_layout, use_container_width=True) csv_bytes = df_layout.to_csv(index=False).encode("utf-8") st.download_button( label="⬇️ Download CSV", data=csv_bytes, file_name=f"kara_layout_sys{system_id}.csv", mime="text/csv", ) st.subheader("Quick Preview (Top 10 Rows)") st.write( df_layout[["Label", "X", "Y", "Z", "Ver_deg"]] .head(10) .style.format({"X": "{:.2f}", "Y": "{:.2f}", "Z": "{:.2f}", "Ver_deg": "{:.1f}"}) ) else: st.info("Set your room + array parameters in the sidebar, then click **Generate Layout (ML Model)**.")