import pandas as pd
import numpy as np
from scipy.sparse import csr_matrix, coo_matrix
import streamlit as st

# multi_project_matching
def calc_matches(filtered_df, project_df, similarity_matrix, top_x):
    st.write(filtered_df.shape)
    st.write(project_df.shape)
    st.write(similarity_matrix.shape)

    # Ensure the matrix is in a suitable format for manipulation
    if not isinstance(similarity_matrix, csr_matrix):
        similarity_matrix = csr_matrix(similarity_matrix)

    # Get indices from dataframes
    filtered_df_indices = filtered_df.index.to_list()
    project_df_indices = project_df.index.to_list()

    # Create mapping dictionaries
    filtered_df_index_map = {i: index for i, index in enumerate(filtered_df_indices)}
    st.write(filtered_df_index_map)
    project_df_index_map = {i: index for i, index in enumerate(project_df_indices)}

    # Select submatrix based on indices from both dataframes
    match_matrix = similarity_matrix[filtered_df_indices, :][:, project_df_indices]

    coo = match_matrix.tocoo()

    data = coo.data
    row_indices = coo.row
    col_indices = coo.col

    top_n = 15
    if len(data) < top_n:
        top_n = len(data) 
    top_n_indices = np.argsort(data)[-top_n:][::-1]

    top_n_percentages = data[top_n_indices]
    top_n_row_indices = row_indices[top_n_indices]
    top_n_col_indices = col_indices[top_n_indices]

    original_row_indices = filtered_df_indices[top_n_row_indices]
    original_col_indices = project_df_indices[top_n_col_indices]

    st.write(match_matrix.shape)

    # Get the linear indices of the top 'top_x' values
    # (flattened index to handle the sparse matrix more effectively)
    #linear_indices = np.argsort(match_matrix.data)[-top_x:]
    #if len(linear_indices) < top_x:
    #    top_x = len(linear_indices)

    # Convert flat indices to 2D indices using the shape of the submatrix
    #top_indices = np.unravel_index(linear_indices, match_matrix.shape)

    # Get the corresponding similarity values
    #top_values = match_matrix.data[linear_indices]

    #flat_data = match_matrix.data

    # Get the indices that would sort the data array in descending order
    #sorted_indices = np.argsort(flat_data)[::-1]

    # Take the first k indices to get the top k maximum values
    #top_indices = sorted_indices[:top_x]
    #top_row_indices = []
    #top_col_indices = []

    #for idx in top_indices:
    #    row, col = np.unravel_index(idx, match_matrix.shape)
    #    top_row_indices.append(row)
    #    top_col_indices.append(col)

    #st.write(top_col_indices)
    # Convert flat indices to 2D row and column indices
    #row_indices, col_indices = match_matrix.nonzero()
    #row_indices = row_indices[top_indices]
    #col_indices = col_indices[top_indices]

    # Get the values corresponding to the top k indices
    #top_values = flat_data[top_indices]


    # Get the values corresponding to the top k indices
    #top_values = match_matrix[row_indices, col_indices]

    #top_filtered_df_indices = [filtered_df_index_map[i] for i in top_col_indices]
    #top_project_df_indices = [project_df_index_map[i] for i in top_row_indices]

    #st.write(top_filtered_df_indices)

    # Create resulting dataframes with top matches and their similarity scores
    p1_df = filtered_df.loc[top_col_indices].copy()
    p1_df['similarity'] = top_values

    p2_df = project_df.loc[top_row_indices].copy()
    p2_df['similarity'] = top_values
    print("finished calc matches")

    return p1_df, p2_df