habitat_classification.py

import pandas as pd

def classify_habitat(species_list, species_data):
    species_to_habitat_weights = species_data.set_index('Species Present').to_dict('index')
    # Initialize habitat votes
    habitat_votes = {col: 0 for col in species_data.columns if col != 'Species Present'}
    # Iterate over each detected species
    # Iterate over each detected species
    unique_species = set(species_list)
    for species in unique_species:
        if species in species_to_habitat_weights:
            # Increment votes based on species presence and its weights
            for habitat in habitat_votes.keys():
                habitat_weight = species_to_habitat_weights[species][habitat]
                habitat_votes[habitat] += habitat_weight

    # Create a DataFrame for the votes
    habitat_votes_df = pd.DataFrame(list(habitat_votes.items()), columns=['Habitat', 'Votes'])

    # Determine the habitat with the highest votes
    best_habitat = max(habitat_votes, key=habitat_votes.get)
    total_votes = sum(habitat_votes.values())
    confidence = habitat_votes[best_habitat] / total_votes if total_votes else 0
    if confidence == 0:
        best_habitat = 'No Classification'

    # Create a DataFrame for the prediction and confidence
    result_df = pd.DataFrame({
        'Prediction': [best_habitat],
        'Confidence': [confidence]
    })

    return result_df, habitat_votes_df

def get_inconclusive():
    result_df = pd.DataFrame({
        'Prediction': ['Inconclusive'],
        'Confidence': 'N/a'})
    probabilities_df = pd.DataFrame({
        'Classification': ['Inconclusive'],
        'Confidence': 'N/a'})
    return result_df, probabilities_df