Summary_v_performance.ipynb

{
 "cells": [
  {
   "cell_type": "code",
   "execution_count": 1,
   "id": "24e2aa4a-e29a-4b4a-ae92-47a8b987e04e",
   "metadata": {},
   "outputs": [
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "[nltk_data] Downloading package punkt to /Users/edmondniu/nltk_data...\n",
      "[nltk_data]   Package punkt is already up-to-date!\n",
      "[nltk_data] Downloading package stopwords to\n",
      "[nltk_data]     /Users/edmondniu/nltk_data...\n",
      "[nltk_data]   Package stopwords is already up-to-date!\n",
      "[nltk_data] Downloading package averaged_perceptron_tagger to\n",
      "[nltk_data]     /Users/edmondniu/nltk_data...\n",
      "[nltk_data]   Package averaged_perceptron_tagger is already up-to-\n",
      "[nltk_data]       date!\n"
     ]
    },
    {
     "data": {
      "text/plain": [
       "True"
      ]
     },
     "execution_count": 1,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "import numpy as np\n",
    "import sklearn as sk\n",
    "import pandas as pd\n",
    "import matplotlib.pyplot as plt\n",
    "import nltk\n",
    "from scipy.optimize import curve_fit\n",
    "from nltk import pos_tag\n",
    "from nltk.tokenize import word_tokenize\n",
    "from nltk.probability import FreqDist\n",
    "from nltk.corpus import stopwords\n",
    "from sklearn.pipeline import make_pipeline\n",
    "from sklearn.linear_model import LinearRegression\n",
    "from sklearn.preprocessing import PolynomialFeatures\n",
    "nltk.download('punkt')\n",
    "nltk.download('stopwords')\n",
    "nltk.download('averaged_perceptron_tagger')"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 2,
   "id": "26fd5dc6-83f8-479e-8202-a6672ead4dd7",
   "metadata": {},
   "outputs": [
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "/var/folders/qq/qm2zshms6rl57mb9s95rh3xc0000gn/T/ipykernel_3691/971401047.py:1: DtypeWarning: Columns (32,33,34,35,36,37,38,39) have mixed types. Specify dtype option on import or set low_memory=False.\n",
      "  sub_responses = pd.read_csv('Data files/responses.csv')\n"
     ]
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "student_id\n",
      "9455ad1a-55b6-44a1-82bc-7d21bc948f5c    2694\n",
      "43cbccae-8a82-401c-a38d-9f43bbc5a163    2663\n",
      "249ad276-075d-4dba-81a2-c375860e29ce    2607\n",
      "7867fc31-3814-4d08-8d68-0336c3df8649    2587\n",
      "843dd9e2-98fd-4fce-9ac2-bcbe7ef7ee88    2586\n",
      "                                        ... \n",
      "2c9302b9-aa06-43b5-a39c-69405cd49b45       5\n",
      "d6f2bf22-a116-4212-a657-8aa70e4ad866       4\n",
      "7a26ed43-861a-41e6-8193-0bc925ab3c5a       4\n",
      "77eb2c8b-cd76-4f9d-9e66-9a8662cb1753       3\n",
      "6d566358-f02f-436c-aba1-e0afd3b5bdd7       2\n",
      "Name: count, Length: 1625, dtype: int64\n"
     ]
    }
   ],
   "source": [
    "sub_responses = pd.read_csv('Data files/responses.csv')\n",
    "sub_checkpoints = pd.read_csv('Data files/checkpoints.csv')\n",
    "check_df = pd.DataFrame(sub_checkpoints)\n",
    "df = pd.DataFrame(sub_responses)\n",
    "counts = df[\"student_id\"].value_counts()\n",
    "print(counts)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 3,
   "id": "3a29191b-6776-42eb-aae1-585a34acaea5",
   "metadata": {},
   "outputs": [],
   "source": [
    "df[\"prompt\"] = df[\"prompt\"].fillna(\"\")\n",
    "filtered_df = df[df[\"prompt\"].str.contains(\"summarize\")]\n",
    "filtered_df = filtered_df[[\"prompt\",\"response\", \"student_id\", \"chapter_number\"]]"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 4,
   "id": "c003b455-90da-477d-a5a6-941e6915cd6f",
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/html": [
       "<div>\n",
       "<style scoped>\n",
       "    .dataframe tbody tr th:only-of-type {\n",
       "        vertical-align: middle;\n",
       "    }\n",
       "\n",
       "    .dataframe tbody tr th {\n",
       "        vertical-align: top;\n",
       "    }\n",
       "\n",
       "    .dataframe thead th {\n",
       "        text-align: right;\n",
       "    }\n",
       "</style>\n",
       "<table border=\"1\" class=\"dataframe\">\n",
       "  <thead>\n",
       "    <tr style=\"text-align: right;\">\n",
       "      <th></th>\n",
       "      <th>prompt</th>\n",
       "      <th>response</th>\n",
       "      <th>student_id</th>\n",
       "      <th>chapter_number</th>\n",
       "    </tr>\n",
       "  </thead>\n",
       "  <tbody>\n",
       "    <tr>\n",
       "      <th>6348</th>\n",
       "      <td>In a paragraph, summarize the main idea(s) in ...</td>\n",
       "      <td>This chapter taught us the goals of this class...</td>\n",
       "      <td>83df1b22-c83a-4412-8e27-1c113bd17907</td>\n",
       "      <td>1.0</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>6349</th>\n",
       "      <td>In a paragraph, summarize the main idea(s) in ...</td>\n",
       "      <td>Statistical model\\nStatistics is profound by s...</td>\n",
       "      <td>dee74669-b1ed-41df-bf06-177c5766071d</td>\n",
       "      <td>1.0</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>6350</th>\n",
       "      <td>In a paragraph, summarize the main idea(s) in ...</td>\n",
       "      <td>The thing to remember when coding in general i...</td>\n",
       "      <td>1d4a6d2c-0cb1-4d19-9218-b673982e9f2f</td>\n",
       "      <td>1.0</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>6351</th>\n",
       "      <td>In a paragraph, summarize the main idea(s) in ...</td>\n",
       "      <td>introduction to the course and statistics! int...</td>\n",
       "      <td>60ecb44a-0a36-4702-9cf4-1f390aba2dca</td>\n",
       "      <td>1.0</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>6352</th>\n",
       "      <td>In a paragraph, summarize the main idea(s) in ...</td>\n",
       "      <td>The main ideas of this chapter are that unders...</td>\n",
       "      <td>c60853dd-e235-477c-a811-be1c3606c7fa</td>\n",
       "      <td>1.0</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>...</th>\n",
       "      <td>...</td>\n",
       "      <td>...</td>\n",
       "      <td>...</td>\n",
       "      <td>...</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>1582918</th>\n",
       "      <td>To summarize, which SS goes with which distances?</td>\n",
       "      <td>[\"From data to complex model's prediction\", \"F...</td>\n",
       "      <td>faecb7cb-6b15-42fc-b0fb-ef2b512c751f</td>\n",
       "      <td>9.0</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>1582941</th>\n",
       "      <td>To summarize, which SS goes with which distances?</td>\n",
       "      <td>[\"From data to complex model's prediction\", \"F...</td>\n",
       "      <td>1784ab66-a261-4426-ba33-46c772d1449c</td>\n",
       "      <td>9.0</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>1582943</th>\n",
       "      <td>To summarize, which SS goes with which distances?</td>\n",
       "      <td>[\"From data to empty model's prediction\", \"Fro...</td>\n",
       "      <td>920a0c75-6a6c-4d60-8706-3f2bff322800</td>\n",
       "      <td>9.0</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>1582954</th>\n",
       "      <td>To summarize, which SS goes with which distances?</td>\n",
       "      <td>[\"From complex model's prediction to empty mod...</td>\n",
       "      <td>ab375b3d-ef49-4a55-af6c-92b587fda293</td>\n",
       "      <td>9.0</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>1582961</th>\n",
       "      <td>To summarize, which SS goes with which distances?</td>\n",
       "      <td>[\"From data to empty model's prediction\", \"Fro...</td>\n",
       "      <td>2b0e5422-d7ef-4161-9abe-3bc0f675e178</td>\n",
       "      <td>9.0</td>\n",
       "    </tr>\n",
       "  </tbody>\n",
       "</table>\n",
       "<p>10385 rows × 4 columns</p>\n",
       "</div>"
      ],
      "text/plain": [
       "                                                    prompt  \\\n",
       "6348     In a paragraph, summarize the main idea(s) in ...   \n",
       "6349     In a paragraph, summarize the main idea(s) in ...   \n",
       "6350     In a paragraph, summarize the main idea(s) in ...   \n",
       "6351     In a paragraph, summarize the main idea(s) in ...   \n",
       "6352     In a paragraph, summarize the main idea(s) in ...   \n",
       "...                                                    ...   \n",
       "1582918  To summarize, which SS goes with which distances?   \n",
       "1582941  To summarize, which SS goes with which distances?   \n",
       "1582943  To summarize, which SS goes with which distances?   \n",
       "1582954  To summarize, which SS goes with which distances?   \n",
       "1582961  To summarize, which SS goes with which distances?   \n",
       "\n",
       "                                                  response  \\\n",
       "6348     This chapter taught us the goals of this class...   \n",
       "6349     Statistical model\\nStatistics is profound by s...   \n",
       "6350     The thing to remember when coding in general i...   \n",
       "6351     introduction to the course and statistics! int...   \n",
       "6352     The main ideas of this chapter are that unders...   \n",
       "...                                                    ...   \n",
       "1582918  [\"From data to complex model's prediction\", \"F...   \n",
       "1582941  [\"From data to complex model's prediction\", \"F...   \n",
       "1582943  [\"From data to empty model's prediction\", \"Fro...   \n",
       "1582954  [\"From complex model's prediction to empty mod...   \n",
       "1582961  [\"From data to empty model's prediction\", \"Fro...   \n",
       "\n",
       "                                   student_id  chapter_number  \n",
       "6348     83df1b22-c83a-4412-8e27-1c113bd17907             1.0  \n",
       "6349     dee74669-b1ed-41df-bf06-177c5766071d             1.0  \n",
       "6350     1d4a6d2c-0cb1-4d19-9218-b673982e9f2f             1.0  \n",
       "6351     60ecb44a-0a36-4702-9cf4-1f390aba2dca             1.0  \n",
       "6352     c60853dd-e235-477c-a811-be1c3606c7fa             1.0  \n",
       "...                                       ...             ...  \n",
       "1582918  faecb7cb-6b15-42fc-b0fb-ef2b512c751f             9.0  \n",
       "1582941  1784ab66-a261-4426-ba33-46c772d1449c             9.0  \n",
       "1582943  920a0c75-6a6c-4d60-8706-3f2bff322800             9.0  \n",
       "1582954  ab375b3d-ef49-4a55-af6c-92b587fda293             9.0  \n",
       "1582961  2b0e5422-d7ef-4161-9abe-3bc0f675e178             9.0  \n",
       "\n",
       "[10385 rows x 4 columns]"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    },
    {
     "data": {
      "text/plain": [
       "(10385, 4)"
      ]
     },
     "execution_count": 4,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "display(filtered_df)\n",
    "filtered_df.shape"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 5,
   "id": "cd7067d2-6f9b-4b35-ad6a-d9a9171e39bb",
   "metadata": {},
   "outputs": [],
   "source": [
    "filtered_df[\"words\"] = filtered_df[\"response\"].str.split()\n",
    "filtered_df[\"length\"] = filtered_df[\"words\"].apply(lambda x: len(x) if isinstance(x, list) else 0)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 6,
   "id": "212008ec-75f5-4b52-8e83-c10603c03b12",
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/html": [
       "<div>\n",
       "<style scoped>\n",
       "    .dataframe tbody tr th:only-of-type {\n",
       "        vertical-align: middle;\n",
       "    }\n",
       "\n",
       "    .dataframe tbody tr th {\n",
       "        vertical-align: top;\n",
       "    }\n",
       "\n",
       "    .dataframe thead th {\n",
       "        text-align: right;\n",
       "    }\n",
       "</style>\n",
       "<table border=\"1\" class=\"dataframe\">\n",
       "  <thead>\n",
       "    <tr style=\"text-align: right;\">\n",
       "      <th></th>\n",
       "      <th>prompt</th>\n",
       "      <th>response</th>\n",
       "      <th>student_id</th>\n",
       "      <th>chapter_number</th>\n",
       "      <th>words</th>\n",
       "      <th>length</th>\n",
       "    </tr>\n",
       "  </thead>\n",
       "  <tbody>\n",
       "    <tr>\n",
       "      <th>6348</th>\n",
       "      <td>In a paragraph, summarize the main idea(s) in ...</td>\n",
       "      <td>This chapter taught us the goals of this class...</td>\n",
       "      <td>83df1b22-c83a-4412-8e27-1c113bd17907</td>\n",
       "      <td>1.0</td>\n",
       "      <td>[This, chapter, taught, us, the, goals, of, th...</td>\n",
       "      <td>55</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>6349</th>\n",
       "      <td>In a paragraph, summarize the main idea(s) in ...</td>\n",
       "      <td>Statistical model\\nStatistics is profound by s...</td>\n",
       "      <td>dee74669-b1ed-41df-bf06-177c5766071d</td>\n",
       "      <td>1.0</td>\n",
       "      <td>[Statistical, model, Statistics, is, profound,...</td>\n",
       "      <td>21</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>6350</th>\n",
       "      <td>In a paragraph, summarize the main idea(s) in ...</td>\n",
       "      <td>The thing to remember when coding in general i...</td>\n",
       "      <td>1d4a6d2c-0cb1-4d19-9218-b673982e9f2f</td>\n",
       "      <td>1.0</td>\n",
       "      <td>[The, thing, to, remember, when, coding, in, g...</td>\n",
       "      <td>179</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>6351</th>\n",
       "      <td>In a paragraph, summarize the main idea(s) in ...</td>\n",
       "      <td>introduction to the course and statistics! int...</td>\n",
       "      <td>60ecb44a-0a36-4702-9cf4-1f390aba2dca</td>\n",
       "      <td>1.0</td>\n",
       "      <td>[introduction, to, the, course, and, statistic...</td>\n",
       "      <td>22</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>6352</th>\n",
       "      <td>In a paragraph, summarize the main idea(s) in ...</td>\n",
       "      <td>The main ideas of this chapter are that unders...</td>\n",
       "      <td>c60853dd-e235-477c-a811-be1c3606c7fa</td>\n",
       "      <td>1.0</td>\n",
       "      <td>[The, main, ideas, of, this, chapter, are, tha...</td>\n",
       "      <td>90</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>...</th>\n",
       "      <td>...</td>\n",
       "      <td>...</td>\n",
       "      <td>...</td>\n",
       "      <td>...</td>\n",
       "      <td>...</td>\n",
       "      <td>...</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>1582918</th>\n",
       "      <td>To summarize, which SS goes with which distances?</td>\n",
       "      <td>[\"From data to complex model's prediction\", \"F...</td>\n",
       "      <td>faecb7cb-6b15-42fc-b0fb-ef2b512c751f</td>\n",
       "      <td>9.0</td>\n",
       "      <td>[[\"From, data, to, complex, model's, predictio...</td>\n",
       "      <td>20</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>1582941</th>\n",
       "      <td>To summarize, which SS goes with which distances?</td>\n",
       "      <td>[\"From data to complex model's prediction\", \"F...</td>\n",
       "      <td>1784ab66-a261-4426-ba33-46c772d1449c</td>\n",
       "      <td>9.0</td>\n",
       "      <td>[[\"From, data, to, complex, model's, predictio...</td>\n",
       "      <td>20</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>1582943</th>\n",
       "      <td>To summarize, which SS goes with which distances?</td>\n",
       "      <td>[\"From data to empty model's prediction\", \"Fro...</td>\n",
       "      <td>920a0c75-6a6c-4d60-8706-3f2bff322800</td>\n",
       "      <td>9.0</td>\n",
       "      <td>[[\"From, data, to, empty, model's, prediction\"...</td>\n",
       "      <td>20</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>1582954</th>\n",
       "      <td>To summarize, which SS goes with which distances?</td>\n",
       "      <td>[\"From complex model's prediction to empty mod...</td>\n",
       "      <td>ab375b3d-ef49-4a55-af6c-92b587fda293</td>\n",
       "      <td>9.0</td>\n",
       "      <td>[[\"From, complex, model's, prediction, to, emp...</td>\n",
       "      <td>20</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>1582961</th>\n",
       "      <td>To summarize, which SS goes with which distances?</td>\n",
       "      <td>[\"From data to empty model's prediction\", \"Fro...</td>\n",
       "      <td>2b0e5422-d7ef-4161-9abe-3bc0f675e178</td>\n",
       "      <td>9.0</td>\n",
       "      <td>[[\"From, data, to, empty, model's, prediction\"...</td>\n",
       "      <td>20</td>\n",
       "    </tr>\n",
       "  </tbody>\n",
       "</table>\n",
       "<p>10385 rows × 6 columns</p>\n",
       "</div>"
      ],
      "text/plain": [
       "                                                    prompt  \\\n",
       "6348     In a paragraph, summarize the main idea(s) in ...   \n",
       "6349     In a paragraph, summarize the main idea(s) in ...   \n",
       "6350     In a paragraph, summarize the main idea(s) in ...   \n",
       "6351     In a paragraph, summarize the main idea(s) in ...   \n",
       "6352     In a paragraph, summarize the main idea(s) in ...   \n",
       "...                                                    ...   \n",
       "1582918  To summarize, which SS goes with which distances?   \n",
       "1582941  To summarize, which SS goes with which distances?   \n",
       "1582943  To summarize, which SS goes with which distances?   \n",
       "1582954  To summarize, which SS goes with which distances?   \n",
       "1582961  To summarize, which SS goes with which distances?   \n",
       "\n",
       "                                                  response  \\\n",
       "6348     This chapter taught us the goals of this class...   \n",
       "6349     Statistical model\\nStatistics is profound by s...   \n",
       "6350     The thing to remember when coding in general i...   \n",
       "6351     introduction to the course and statistics! int...   \n",
       "6352     The main ideas of this chapter are that unders...   \n",
       "...                                                    ...   \n",
       "1582918  [\"From data to complex model's prediction\", \"F...   \n",
       "1582941  [\"From data to complex model's prediction\", \"F...   \n",
       "1582943  [\"From data to empty model's prediction\", \"Fro...   \n",
       "1582954  [\"From complex model's prediction to empty mod...   \n",
       "1582961  [\"From data to empty model's prediction\", \"Fro...   \n",
       "\n",
       "                                   student_id  chapter_number  \\\n",
       "6348     83df1b22-c83a-4412-8e27-1c113bd17907             1.0   \n",
       "6349     dee74669-b1ed-41df-bf06-177c5766071d             1.0   \n",
       "6350     1d4a6d2c-0cb1-4d19-9218-b673982e9f2f             1.0   \n",
       "6351     60ecb44a-0a36-4702-9cf4-1f390aba2dca             1.0   \n",
       "6352     c60853dd-e235-477c-a811-be1c3606c7fa             1.0   \n",
       "...                                       ...             ...   \n",
       "1582918  faecb7cb-6b15-42fc-b0fb-ef2b512c751f             9.0   \n",
       "1582941  1784ab66-a261-4426-ba33-46c772d1449c             9.0   \n",
       "1582943  920a0c75-6a6c-4d60-8706-3f2bff322800             9.0   \n",
       "1582954  ab375b3d-ef49-4a55-af6c-92b587fda293             9.0   \n",
       "1582961  2b0e5422-d7ef-4161-9abe-3bc0f675e178             9.0   \n",
       "\n",
       "                                                     words  length  \n",
       "6348     [This, chapter, taught, us, the, goals, of, th...      55  \n",
       "6349     [Statistical, model, Statistics, is, profound,...      21  \n",
       "6350     [The, thing, to, remember, when, coding, in, g...     179  \n",
       "6351     [introduction, to, the, course, and, statistic...      22  \n",
       "6352     [The, main, ideas, of, this, chapter, are, tha...      90  \n",
       "...                                                    ...     ...  \n",
       "1582918  [[\"From, data, to, complex, model's, predictio...      20  \n",
       "1582941  [[\"From, data, to, complex, model's, predictio...      20  \n",
       "1582943  [[\"From, data, to, empty, model's, prediction\"...      20  \n",
       "1582954  [[\"From, complex, model's, prediction, to, emp...      20  \n",
       "1582961  [[\"From, data, to, empty, model's, prediction\"...      20  \n",
       "\n",
       "[10385 rows x 6 columns]"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "<class 'pandas.core.frame.DataFrame'>\n",
      "Index: 10385 entries, 6348 to 1582961\n",
      "Data columns (total 6 columns):\n",
      " #   Column          Non-Null Count  Dtype  \n",
      "---  ------          --------------  -----  \n",
      " 0   prompt          10385 non-null  object \n",
      " 1   response        10356 non-null  object \n",
      " 2   student_id      10385 non-null  object \n",
      " 3   chapter_number  10385 non-null  float64\n",
      " 4   words           10356 non-null  object \n",
      " 5   length          10385 non-null  int64  \n",
      "dtypes: float64(1), int64(1), object(4)\n",
      "memory usage: 567.9+ KB\n"
     ]
    }
   ],
   "source": [
    "display(filtered_df)\n",
    "filtered_df.info()"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 7,
   "id": "bad432ad-c38c-4849-a70f-64d7dbf3be2b",
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "student_id                            chapter_number\n",
      "001824fb-a2fd-431d-aef6-7a1250d97a62  1.0               131\n",
      "                                      2.0                91\n",
      "                                      3.0                65\n",
      "                                      5.0                71\n",
      "                                      6.0                73\n",
      "Name: length, dtype: int64\n"
     ]
    },
    {
     "data": {
      "text/plain": [
       "pandas.core.series.Series"
      ]
     },
     "execution_count": 7,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "length_summaries = filtered_df.groupby(['student_id', 'chapter_number'])['length'].sum()\n",
    "print(length_summaries.head())\n",
    "type(length_summaries)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 8,
   "id": "8664ff30-5f31-4d7f-8b40-53043d8d683d",
   "metadata": {},
   "outputs": [],
   "source": [
    "#export data\n",
    "\n",
    "import os\n",
    "subfolder = 'random subset of data files'\n",
    "filename = 'length_series.csv'\n",
    "filepath = os.path.join(subfolder, filename)\n",
    "length_summaries.to_csv(filepath)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 9,
   "id": "c8003178-d8fe-4da0-bb5c-d3fe76544898",
   "metadata": {},
   "outputs": [],
   "source": [
    "def calculate_lexical_density(text):\n",
    "    # Tokenize the text into words\n",
    "    words = word_tokenize(text)\n",
    "\n",
    "    # Remove stopwords\n",
    "    stop_words = set(stopwords.words('english'))\n",
    "    filtered_words = [word.lower() for word in words if word.isalpha() and word.lower() not in stop_words]\n",
    "\n",
    "    # Calculate lexical density (normalized)\n",
    "    unique_words = set(filtered_words)\n",
    "    \n",
    "    # Handle the case of one-word texts\n",
    "    if len(words) <= 5:\n",
    "        lexical_density = 0\n",
    "    else:\n",
    "        lexical_density = len(unique_words) / len(words) if len(words) > 0 else 0\n",
    "\n",
    "    return lexical_density\n",
    "\n",
    "def calc_grammatical_score(text):\n",
    "    tokens = word_tokenize(text)\n",
    "    pos_tags = pos_tag(tokens)\n",
    "\n",
    "    # Calculate part-of-speech diversity\n",
    "    pos_set = set(tag for word, tag in pos_tags)\n",
    "    pos_diversity = len(pos_set) / len(tokens)\n",
    "    return pos_diversity\n",
    "    "
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 10,
   "id": "44a4ca5c-2bda-48cd-82a7-cfcf17993fcd",
   "metadata": {},
   "outputs": [],
   "source": [
    "filtered_df[\"density_score\"] = filtered_df[\"response\"].apply(lambda x: calculate_lexical_density(x) if isinstance(x, str) else 0)\n",
    "filtered_df[\"grammatical_score\"] = filtered_df[\"response\"].apply(lambda x: calc_grammatical_score(x) if isinstance(x, str) else 0)\n",
    "\n",
    "average_density = filtered_df.groupby('student_id')['density_score'].apply(lambda x: x.mean() if x.dtype == 'float64' else 0).reset_index(name='average_density')\n",
    "average_length = filtered_df.groupby('student_id')['length'].apply(lambda x: x.mean() if x.dtype == 'int64' else 0).reset_index(name='average_length')\n",
    "average_gram = filtered_df.groupby('student_id')['grammatical_score'].apply(lambda x: x.mean() if x.dtype == 'float64' else 0).reset_index(name='average_gram')"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 11,
   "id": "d060a12f-ce1f-48f3-90e3-2651a89d2048",
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "                                student_id  average_length\n",
      "0     001824fb-a2fd-431d-aef6-7a1250d97a62       97.100000\n",
      "1     0096491a-77bf-4e79-b906-fc7a8e5e57a6      152.900000\n",
      "2     00a53a52-aa0e-4ddc-afa8-2c50c21dfd03       36.333333\n",
      "3     00fe5bb2-6b64-4a36-a706-eac4dc124ea7       82.300000\n",
      "4     016e2fcf-0cc2-42ad-844d-e1057b5081cc      210.000000\n",
      "...                                    ...             ...\n",
      "1481  fecc783d-1cd1-4940-aa16-695c4652a536      147.000000\n",
      "1482  fecf18d0-a58a-4d11-a353-ea73fbfa6860       47.111111\n",
      "1483  fef083e8-7167-4f42-82d5-d2e893aa408a       30.666667\n",
      "1484  ff55bab4-812b-405b-9856-8bedf9fb11f9       90.300000\n",
      "1485  ff6ec9fe-de1d-4b45-8136-59465d9c85ab       45.000000\n",
      "\n",
      "[1486 rows x 2 columns]\n"
     ]
    }
   ],
   "source": [
    "print(average_length)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 12,
   "id": "b092f3f4-3234-4e43-8e88-c2ff2baaec28",
   "metadata": {},
   "outputs": [],
   "source": [
    "x = average_density[\"average_density\"]\n",
    "average_score = check_df.groupby('student_id')['n_correct'].apply(lambda x: x.mean() if x.dtype == 'float64' else 0).reset_index(name='average_score')\n",
    "y = average_score[\"average_score\"]\n"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 13,
   "id": "54a4e6ae-41f4-4178-b462-f677b059d9dc",
   "metadata": {},
   "outputs": [],
   "source": [
    "result = pd.merge(average_score, average_density, on='student_id', how='inner').fillna(0)\n",
    "x = result[\"average_density\"]\n",
    "y = result[\"average_score\"]"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 14,
   "id": "097803d4-c989-4284-a921-dbc04a4d304b",
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "<matplotlib.collections.PathCollection at 0x2919eba10>"
      ]
     },
     "execution_count": 14,
     "metadata": {},
     "output_type": "execute_result"
    },
    {
     "data": {
      "image/png": "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",
      "text/plain": [
       "<Figure size 640x480 with 1 Axes>"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    }
   ],
   "source": [
    "plt.scatter(x, y)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 15,
   "id": "0232dc60-2948-495f-835c-5b16b1600627",
   "metadata": {},
   "outputs": [
    {
     "ename": "NameError",
     "evalue": "name 'poly_model' is not defined",
     "output_type": "error",
     "traceback": [
      "\u001b[0;31m---------------------------------------------------------------------------\u001b[0m",
      "\u001b[0;31mNameError\u001b[0m                                 Traceback (most recent call last)",
      "Cell \u001b[0;32mIn[15], line 2\u001b[0m\n\u001b[1;32m      1\u001b[0m \u001b[38;5;66;03m# Predict on the original data\u001b[39;00m\n\u001b[0;32m----> 2\u001b[0m y_pred \u001b[38;5;241m=\u001b[39m \u001b[43mpoly_model\u001b[49m\u001b[38;5;241m.\u001b[39mpredict(z[:, np\u001b[38;5;241m.\u001b[39mnewaxis])\n\u001b[1;32m      4\u001b[0m \u001b[38;5;66;03m# Plot the polynomial fit\u001b[39;00m\n\u001b[1;32m      5\u001b[0m plt\u001b[38;5;241m.\u001b[39mplot(z, y_pred, color\u001b[38;5;241m=\u001b[39m\u001b[38;5;124m'\u001b[39m\u001b[38;5;124mred\u001b[39m\u001b[38;5;124m'\u001b[39m, label\u001b[38;5;241m=\u001b[39m\u001b[38;5;124mf\u001b[39m\u001b[38;5;124m'\u001b[39m\u001b[38;5;124mPolynomial Fit (Degree \u001b[39m\u001b[38;5;132;01m{\u001b[39;00mdegree\u001b[38;5;132;01m}\u001b[39;00m\u001b[38;5;124m)\u001b[39m\u001b[38;5;124m'\u001b[39m)\n",
      "\u001b[0;31mNameError\u001b[0m: name 'poly_model' is not defined"
     ]
    }
   ],
   "source": [
    "# Predict on the original data\n",
    "y_pred = poly_model.predict(z[:, np.newaxis])\n",
    "\n",
    "# Plot the polynomial fit\n",
    "plt.plot(z, y_pred, color='red', label=f'Polynomial Fit (Degree {degree})')\n",
    "\n",
    "# Add labels and legend\n",
    "plt.xlabel('Average Length')\n",
    "plt.ylabel('Average Score')\n",
    "plt.legend()\n",
    "\n",
    "# Show the plot\n",
    "plt.show()\n",
    "\n",
    "# Show the plot\n",
    "plt.show()"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "id": "1dd2358f-db59-41c9-8248-baf7f8aadef4",
   "metadata": {},
   "outputs": [],
   "source": [
    "result1 = pd.merge(average_score, average_gram, on='student_id', how='inner').fillna(0)\n",
    "a = result1[\"average_gram\"]\n",
    "y = result1[\"average_score\"]\n",
    "plt.scatter(a, y)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "id": "d14342cc-6633-4654-9896-5a124ff0d8f3",
   "metadata": {},
   "outputs": [],
   "source": []
  }
 ],
 "metadata": {
  "kernelspec": {
   "display_name": "Python 3 (ipykernel)",
   "language": "python",
   "name": "python3"
  },
  "language_info": {
   "codemirror_mode": {
    "name": "ipython",
    "version": 3
   },
   "file_extension": ".py",
   "mimetype": "text/x-python",
   "name": "python",
   "nbconvert_exporter": "python",
   "pygments_lexer": "ipython3",
   "version": "3.11.5"
  }
 },
 "nbformat": 4,
 "nbformat_minor": 5
}