create_exp2_outputs.py

import openai
import pandas as pd
import pandas as pd
import json
import urllib
import math
import time
import random
import re
from tqdm import tqdm
from io import StringIO 
import exp_lib



def experiment_2_trial(data_df, model_name):
  x = data_df.sample(frac=1)
  train_df = x.drop_duplicates('Misconception ID')
  test_df = x.iloc[::-1].drop_duplicates('Misconception ID')
  test_df = test_df.reset_index()
  topics = [
       'Ratios and proportional reasoning', 
       'Number Operations',
       'Patterns, relationships, and functions', 
       'Number sense',
       'Algebraic representations',
       'Variables, expressions, and operations',
       'Equations and inequalities',
       'Properties of number and operations'
       ]
  # now, iterate by topic and slice each topic data out of train_df, test_df
  topic_test_dfs = []
  for topic in topics:
    topic_test_df = test_df[test_df['Topic'] == topic].copy()
    topic_test_df = topic_test_df.reset_index()
    topic_train_df = train_df[train_df['Topic'] == topic].copy()
    prompt = exp_lib.generate_prompt_test_batch(topic_train_df.to_dict(orient='records'), topic_test_df.to_dict(orient='records'))
    response = exp_lib.get_gpt4_diagnosis(model_name, prompt)
    response_df = pd.read_csv(StringIO(response), header=None, names=["test_example", "diagnosis"])
    topic_test_df["Predicted Diagnosis"] = response_df["diagnosis"].str.strip()
    topic_test_df["Model"] = model_name
    topic_test_dfs.append(topic_test_df)

  topic_test_df2 = pd.concat(topic_test_dfs)  
  return topic_test_df2[['Misconception ID', 'Example Number', 'Topic', 'Predicted Diagnosis', 'Model']]


def experiment_2(input_file_path, model_name, num_iterations, output_file_path):
  data_df = pd.read_json(input_file_path)
  experiment_2_results_list = []
  for i in tqdm(range(num_iterations)):
    try:
      trial_result = experiment_2_trial(data_df, model_name)
      trial_result['Trial'] = i 
      experiment_2_results_list.append(trial_result)
    except Exception as e:
      print(e)  
  experiment_2_results_df = pd.concat(experiment_2_results_list)
  experiment_2_results_df['Correct'] = (experiment_2_results_df['Misconception ID'] == experiment_2_results_df['Predicted Diagnosis'])
  experiment_2_results_df.to_csv(output_file_path)


if __name__ == '__main__':
  experiment_name = 'experiment_2'
  input_file_path = 'data/data.json' 
  model_name = 'gpt-4-turbo'
  num_iterations = 100
  output_file_path = f'outputs/{experiment_name}_{model_name}_{num_iterations}iters.csv'

  experiment_2(
    input_file_path,
    model_name,
    num_iterations,
    output_file_path
  )