fit.ipynb

{
 "cells": [
  {
   "cell_type": "code",
   "execution_count": 2,
   "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>Model Category</th>\n",
       "      <th>Model</th>\n",
       "      <th>ModelAbv</th>\n",
       "      <th>Variant</th>\n",
       "      <th>Parameter Name</th>\n",
       "      <th>Parameter Type</th>\n",
       "      <th>Description</th>\n",
       "      <th>Prior Distribution</th>\n",
       "      <th>Default Prior Distribution</th>\n",
       "      <th>AgentClass</th>\n",
       "      <th>SHORTCODE</th>\n",
       "      <th>FittingExtras</th>\n",
       "      <th>OptimizationAlgorithm</th>\n",
       "      <th>FitDir</th>\n",
       "      <th>RPE</th>\n",
       "      <th>Forgetting</th>\n",
       "      <th>Future Discounting</th>\n",
       "      <th>Perseverance</th>\n",
       "      <th>On-Policy</th>\n",
       "    </tr>\n",
       "  </thead>\n",
       "  <tbody>\n",
       "    <tr>\n",
       "      <th>7</th>\n",
       "      <td>Value-Based RL</td>\n",
       "      <td>RPE-free Q-Learning</td>\n",
       "      <td>RF-QL</td>\n",
       "      <td>acceptreject</td>\n",
       "      <td>alpha</td>\n",
       "      <td>learningrule</td>\n",
       "      <td>Learning Rate</td>\n",
       "      <td>Beta</td>\n",
       "      <td>Beta(1,1)</td>\n",
       "      <td>RewardLearner_acceptreject</td>\n",
       "      <td>RBR</td>\n",
       "      <td>None</td>\n",
       "      <td>NUTS</td>\n",
       "      <td>modelfits/acceptreject/RBR_2022_04_09_19_31_45.nc</td>\n",
       "      <td>-</td>\n",
       "      <td>-</td>\n",
       "      <td>-</td>\n",
       "      <td>-</td>\n",
       "      <td>-</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>8</th>\n",
       "      <td>Value-Based RL</td>\n",
       "      <td>RPE-free Q-Learning</td>\n",
       "      <td>RF-QL</td>\n",
       "      <td>acceptreject</td>\n",
       "      <td>weight</td>\n",
       "      <td>policy</td>\n",
       "      <td>Q-value weight</td>\n",
       "      <td>Normal</td>\n",
       "      <td>Normal(0,1)</td>\n",
       "      <td>RewardLearner_acceptreject</td>\n",
       "      <td>RBR</td>\n",
       "      <td>None</td>\n",
       "      <td>NUTS</td>\n",
       "      <td>modelfits/acceptreject/RBR_2022_04_09_19_31_45.nc</td>\n",
       "      <td>-</td>\n",
       "      <td>-</td>\n",
       "      <td>-</td>\n",
       "      <td>-</td>\n",
       "      <td>-</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>9</th>\n",
       "      <td>Value-Based RL</td>\n",
       "      <td>RPE-free Q-Learning</td>\n",
       "      <td>RF-QL</td>\n",
       "      <td>acceptreject</td>\n",
       "      <td>intercept</td>\n",
       "      <td>policy</td>\n",
       "      <td>Q-value intercept</td>\n",
       "      <td>Normal</td>\n",
       "      <td>Normal(0,1)</td>\n",
       "      <td>RewardLearner_acceptreject</td>\n",
       "      <td>RBR</td>\n",
       "      <td>None</td>\n",
       "      <td>NUTS</td>\n",
       "      <td>modelfits/acceptreject/RBR_2022_04_09_19_31_45.nc</td>\n",
       "      <td>-</td>\n",
       "      <td>-</td>\n",
       "      <td>-</td>\n",
       "      <td>-</td>\n",
       "      <td>-</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>17</th>\n",
       "      <td>Value-Based RL</td>\n",
       "      <td>Forgetting RPE-free Q-Learning</td>\n",
       "      <td>F-RF-QL</td>\n",
       "      <td>acceptreject</td>\n",
       "      <td>alpha</td>\n",
       "      <td>learningrule</td>\n",
       "      <td>Learning Rate</td>\n",
       "      <td>Beta</td>\n",
       "      <td>Beta(1,1)</td>\n",
       "      <td>ForgettingRewardLearner_acceptreject</td>\n",
       "      <td>FRBR</td>\n",
       "      <td>None</td>\n",
       "      <td>NUTS</td>\n",
       "      <td>modelfits/acceptreject/FRBR_2022_04_22_22_08_2...</td>\n",
       "      <td>-</td>\n",
       "      <td>+</td>\n",
       "      <td>-</td>\n",
       "      <td>-</td>\n",
       "      <td>-</td>\n",
       "    </tr>\n",
       "    <tr>\n",
       "      <th>18</th>\n",
       "      <td>Value-Based RL</td>\n",
       "      <td>Forgetting RPE-free Q-Learning</td>\n",
       "      <td>F-RF-QL</td>\n",
       "      <td>acceptreject</td>\n",
       "      <td>weight</td>\n",
       "      <td>policy</td>\n",
       "      <td>Q-value weight</td>\n",
       "      <td>Normal</td>\n",
       "      <td>Normal(0,1)</td>\n",
       "      <td>ForgettingRewardLearner_acceptreject</td>\n",
       "      <td>FRBR</td>\n",
       "      <td>None</td>\n",
       "      <td>NUTS</td>\n",
       "      <td>modelfits/acceptreject/FRBR_2022_04_22_22_08_2...</td>\n",
       "      <td>-</td>\n",
       "      <td>+</td>\n",
       "      <td>-</td>\n",
       "      <td>-</td>\n",
       "      <td>-</td>\n",
       "    </tr>\n",
       "  </tbody>\n",
       "</table>\n",
       "</div>"
      ],
      "text/plain": [
       "    Model Category                           Model ModelAbv       Variant  \\\n",
       "7   Value-Based RL             RPE-free Q-Learning    RF-QL  acceptreject   \n",
       "8   Value-Based RL             RPE-free Q-Learning    RF-QL  acceptreject   \n",
       "9   Value-Based RL             RPE-free Q-Learning    RF-QL  acceptreject   \n",
       "17  Value-Based RL  Forgetting RPE-free Q-Learning  F-RF-QL  acceptreject   \n",
       "18  Value-Based RL  Forgetting RPE-free Q-Learning  F-RF-QL  acceptreject   \n",
       "\n",
       "   Parameter Name Parameter Type        Description Prior Distribution  \\\n",
       "7           alpha   learningrule      Learning Rate               Beta   \n",
       "8          weight         policy     Q-value weight             Normal   \n",
       "9       intercept         policy  Q-value intercept             Normal   \n",
       "17          alpha   learningrule      Learning Rate               Beta   \n",
       "18         weight         policy     Q-value weight             Normal   \n",
       "\n",
       "   Default Prior Distribution                            AgentClass SHORTCODE  \\\n",
       "7                   Beta(1,1)            RewardLearner_acceptreject       RBR   \n",
       "8                 Normal(0,1)            RewardLearner_acceptreject       RBR   \n",
       "9                 Normal(0,1)            RewardLearner_acceptreject       RBR   \n",
       "17                  Beta(1,1)  ForgettingRewardLearner_acceptreject      FRBR   \n",
       "18                Normal(0,1)  ForgettingRewardLearner_acceptreject      FRBR   \n",
       "\n",
       "   FittingExtras OptimizationAlgorithm  \\\n",
       "7           None                  NUTS   \n",
       "8           None                  NUTS   \n",
       "9           None                  NUTS   \n",
       "17          None                  NUTS   \n",
       "18          None                  NUTS   \n",
       "\n",
       "                                               FitDir RPE Forgetting  \\\n",
       "7   modelfits/acceptreject/RBR_2022_04_09_19_31_45.nc   -          -   \n",
       "8   modelfits/acceptreject/RBR_2022_04_09_19_31_45.nc   -          -   \n",
       "9   modelfits/acceptreject/RBR_2022_04_09_19_31_45.nc   -          -   \n",
       "17  modelfits/acceptreject/FRBR_2022_04_22_22_08_2...   -          +   \n",
       "18  modelfits/acceptreject/FRBR_2022_04_22_22_08_2...   -          +   \n",
       "\n",
       "   Future Discounting Perseverance On-Policy  \n",
       "7                   -            -         -  \n",
       "8                   -            -         -  \n",
       "9                   -            -         -  \n",
       "17                  -            -         -  \n",
       "18                  -            -         -  "
      ]
     },
     "execution_count": 2,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "import pandas as pd\n",
    "\n",
    "model_database = pd.read_csv('model_description.csv')\n",
    "# keep only acceptreject Variant\n",
    "model_database = model_database[model_database['Variant'] == 'acceptreject']\n",
    "model_database.head()"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 3,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "array(['RBR', 'FRBR', 'IQBR', 'CQBR', 'FQBR', 'FCQBR', 'DFQBR', 'DFCQBR',\n",
       "       'DEQBR', 'DECQBR', 'OSQBR', 'OSCQBR', 'FOSQBR', 'FOSCQBR',\n",
       "       'DFOSQBR', 'DFOSCQBR', 'SBR', 'ESBR', 'HQBR', 'FHQBR', 'HCQBR',\n",
       "       'FHCQBR'], dtype=object)"
      ]
     },
     "execution_count": 3,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "# get unique SHORTCODEs\n",
    "model_database['SHORTCODE'].unique()"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 5,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "bsub -n4 -J \"RBRfit\" -o \"RBR.output\" \"python FlYMazeRL/flymazerl/scripts/classical_fitting.py --agent RBR --n_burn 5000 --n_steps 5000 --n_chains 4 --n_cores 4 --target_acceptance_rate 0.95 --save_path /groups/turner/turnerlab/Rishika/FlYMazeRL_Fits/acceptreject/mohanta2022/\"\n",
      "bsub -n4 -J \"FRBRfit\" -o \"FRBR.output\" \"python FlYMazeRL/flymazerl/scripts/classical_fitting.py --agent FRBR --n_burn 5000 --n_steps 5000 --n_chains 4 --n_cores 4 --target_acceptance_rate 0.95 --save_path /groups/turner/turnerlab/Rishika/FlYMazeRL_Fits/acceptreject/mohanta2022/\"\n",
      "bsub -n4 -J \"IQBRfit\" -o \"IQBR.output\" \"python FlYMazeRL/flymazerl/scripts/classical_fitting.py --agent IQBR --n_burn 5000 --n_steps 5000 --n_chains 4 --n_cores 4 --target_acceptance_rate 0.95 --save_path /groups/turner/turnerlab/Rishika/FlYMazeRL_Fits/acceptreject/mohanta2022/\"\n",
      "bsub -n4 -J \"CQBRfit\" -o \"CQBR.output\" \"python FlYMazeRL/flymazerl/scripts/classical_fitting.py --agent CQBR --n_burn 5000 --n_steps 5000 --n_chains 4 --n_cores 4 --target_acceptance_rate 0.95 --save_path /groups/turner/turnerlab/Rishika/FlYMazeRL_Fits/acceptreject/mohanta2022/\"\n",
      "bsub -n4 -J \"FQBRfit\" -o \"FQBR.output\" \"python FlYMazeRL/flymazerl/scripts/classical_fitting.py --agent FQBR --n_burn 5000 --n_steps 5000 --n_chains 4 --n_cores 4 --target_acceptance_rate 0.95 --save_path /groups/turner/turnerlab/Rishika/FlYMazeRL_Fits/acceptreject/mohanta2022/\"\n",
      "bsub -n4 -J \"FCQBRfit\" -o \"FCQBR.output\" \"python FlYMazeRL/flymazerl/scripts/classical_fitting.py --agent FCQBR --n_burn 5000 --n_steps 5000 --n_chains 4 --n_cores 4 --target_acceptance_rate 0.95 --save_path /groups/turner/turnerlab/Rishika/FlYMazeRL_Fits/acceptreject/mohanta2022/\"\n",
      "bsub -n4 -J \"DFQBRfit\" -o \"DFQBR.output\" \"python FlYMazeRL/flymazerl/scripts/classical_fitting.py --agent DFQBR --n_burn 5000 --n_steps 5000 --n_chains 4 --n_cores 4 --target_acceptance_rate 0.95 --save_path /groups/turner/turnerlab/Rishika/FlYMazeRL_Fits/acceptreject/mohanta2022/\"\n",
      "bsub -n4 -J \"DFCQBRfit\" -o \"DFCQBR.output\" \"python FlYMazeRL/flymazerl/scripts/classical_fitting.py --agent DFCQBR --n_burn 5000 --n_steps 5000 --n_chains 4 --n_cores 4 --target_acceptance_rate 0.95 --save_path /groups/turner/turnerlab/Rishika/FlYMazeRL_Fits/acceptreject/mohanta2022/\"\n",
      "bsub -n4 -J \"DEQBRfit\" -o \"DEQBR.output\" \"python FlYMazeRL/flymazerl/scripts/classical_fitting.py --agent DEQBR --n_burn 5000 --n_steps 5000 --n_chains 4 --n_cores 4 --target_acceptance_rate 0.95 --save_path /groups/turner/turnerlab/Rishika/FlYMazeRL_Fits/acceptreject/mohanta2022/\"\n",
      "bsub -n4 -J \"DECQBRfit\" -o \"DECQBR.output\" \"python FlYMazeRL/flymazerl/scripts/classical_fitting.py --agent DECQBR --n_burn 5000 --n_steps 5000 --n_chains 4 --n_cores 4 --target_acceptance_rate 0.95 --save_path /groups/turner/turnerlab/Rishika/FlYMazeRL_Fits/acceptreject/mohanta2022/\"\n",
      "bsub -n4 -J \"OSQBRfit\" -o \"OSQBR.output\" \"python FlYMazeRL/flymazerl/scripts/classical_fitting.py --agent OSQBR --n_burn 5000 --n_steps 5000 --n_chains 4 --n_cores 4 --target_acceptance_rate 0.95 --save_path /groups/turner/turnerlab/Rishika/FlYMazeRL_Fits/acceptreject/mohanta2022/\"\n",
      "bsub -n4 -J \"OSCQBRfit\" -o \"OSCQBR.output\" \"python FlYMazeRL/flymazerl/scripts/classical_fitting.py --agent OSCQBR --n_burn 5000 --n_steps 5000 --n_chains 4 --n_cores 4 --target_acceptance_rate 0.95 --save_path /groups/turner/turnerlab/Rishika/FlYMazeRL_Fits/acceptreject/mohanta2022/\"\n",
      "bsub -n4 -J \"FOSQBRfit\" -o \"FOSQBR.output\" \"python FlYMazeRL/flymazerl/scripts/classical_fitting.py --agent FOSQBR --n_burn 5000 --n_steps 5000 --n_chains 4 --n_cores 4 --target_acceptance_rate 0.95 --save_path /groups/turner/turnerlab/Rishika/FlYMazeRL_Fits/acceptreject/mohanta2022/\"\n",
      "bsub -n4 -J \"FOSCQBRfit\" -o \"FOSCQBR.output\" \"python FlYMazeRL/flymazerl/scripts/classical_fitting.py --agent FOSCQBR --n_burn 5000 --n_steps 5000 --n_chains 4 --n_cores 4 --target_acceptance_rate 0.95 --save_path /groups/turner/turnerlab/Rishika/FlYMazeRL_Fits/acceptreject/mohanta2022/\"\n",
      "bsub -n4 -J \"DFOSQBRfit\" -o \"DFOSQBR.output\" \"python FlYMazeRL/flymazerl/scripts/classical_fitting.py --agent DFOSQBR --n_burn 5000 --n_steps 5000 --n_chains 4 --n_cores 4 --target_acceptance_rate 0.95 --save_path /groups/turner/turnerlab/Rishika/FlYMazeRL_Fits/acceptreject/mohanta2022/\"\n",
      "bsub -n4 -J \"DFOSCQBRfit\" -o \"DFOSCQBR.output\" \"python FlYMazeRL/flymazerl/scripts/classical_fitting.py --agent DFOSCQBR --n_burn 5000 --n_steps 5000 --n_chains 4 --n_cores 4 --target_acceptance_rate 0.95 --save_path /groups/turner/turnerlab/Rishika/FlYMazeRL_Fits/acceptreject/mohanta2022/\"\n",
      "bsub -n4 -J \"SBRfit\" -o \"SBR.output\" \"python FlYMazeRL/flymazerl/scripts/classical_fitting.py --agent SBR --n_burn 5000 --n_steps 5000 --n_chains 4 --n_cores 4 --target_acceptance_rate 0.95 --save_path /groups/turner/turnerlab/Rishika/FlYMazeRL_Fits/acceptreject/mohanta2022/\"\n",
      "bsub -n4 -J \"ESBRfit\" -o \"ESBR.output\" \"python FlYMazeRL/flymazerl/scripts/classical_fitting.py --agent ESBR --n_burn 5000 --n_steps 5000 --n_chains 4 --n_cores 4 --target_acceptance_rate 0.95 --save_path /groups/turner/turnerlab/Rishika/FlYMazeRL_Fits/acceptreject/mohanta2022/\"\n",
      "bsub -n4 -J \"HQBRfit\" -o \"HQBR.output\" \"python FlYMazeRL/flymazerl/scripts/classical_fitting.py --agent HQBR --n_burn 5000 --n_steps 5000 --n_chains 4 --n_cores 4 --target_acceptance_rate 0.95 --save_path /groups/turner/turnerlab/Rishika/FlYMazeRL_Fits/acceptreject/mohanta2022/\"\n",
      "bsub -n4 -J \"FHQBRfit\" -o \"FHQBR.output\" \"python FlYMazeRL/flymazerl/scripts/classical_fitting.py --agent FHQBR --n_burn 5000 --n_steps 5000 --n_chains 4 --n_cores 4 --target_acceptance_rate 0.95 --save_path /groups/turner/turnerlab/Rishika/FlYMazeRL_Fits/acceptreject/mohanta2022/\"\n",
      "bsub -n4 -J \"HCQBRfit\" -o \"HCQBR.output\" \"python FlYMazeRL/flymazerl/scripts/classical_fitting.py --agent HCQBR --n_burn 5000 --n_steps 5000 --n_chains 4 --n_cores 4 --target_acceptance_rate 0.95 --save_path /groups/turner/turnerlab/Rishika/FlYMazeRL_Fits/acceptreject/mohanta2022/\"\n",
      "bsub -n4 -J \"FHCQBRfit\" -o \"FHCQBR.output\" \"python FlYMazeRL/flymazerl/scripts/classical_fitting.py --agent FHCQBR --n_burn 5000 --n_steps 5000 --n_chains 4 --n_cores 4 --target_acceptance_rate 0.95 --save_path /groups/turner/turnerlab/Rishika/FlYMazeRL_Fits/acceptreject/mohanta2022/\"\n"
     ]
    }
   ],
   "source": [
    "format = 'bsub -n4 -J \"{}fit\" -o \"{}.output\" \"python FlYMazeRL/flymazerl/scripts/classical_fitting.py --agent {} --n_burn 5000 --n_steps 5000 --n_chains 4 --n_cores 4 --target_acceptance_rate 0.95 --save_path /groups/turner/turnerlab/Rishika/FlYMazeRL_Fits/acceptreject/mohanta2022/\"'\n",
    "\n",
    "for i in model_database['SHORTCODE'].unique():\n",
    "    print(format.format(i, i, i))"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "metadata": {},
   "outputs": [],
   "source": []
  }
 ],
 "metadata": {
  "kernelspec": {
   "display_name": "Python 3.9.12 ('flymazerl')",
   "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": "3.9.12"
  },
  "orig_nbformat": 4,
  "vscode": {
   "interpreter": {
    "hash": "97727a114667ed15117250b5fec8748dc40cf28e874283bb31e511af3ebb2b4c"
   }
  }
 },
 "nbformat": 4,
 "nbformat_minor": 2
}