Using AgentPy Provide the code proposed for modeling an agent in a room. Consider the following rules:

Environment:

A room is a rectangle, arranged in cells, with m

rows and

n columns.

A cell can contain a reward

$($

negative value

$)$

$$or a penalty

$($

positive value

$)$

$.$

As the robot passes over the cell, it collects and removes the item it contains.

Agent:

A robot can obtain the information of each cell in the room using a coordinate system

$($

row

$,$

$$column

$)$

$.$

The robot takes

$1$

$$unit of time to travel from one cell to any neighboring cell

$($

up

$,$

$$down$,$ right or left

$)$

$.$

The robot spends

$1$

$$unit of energy to travel from one cell to any neighboring cell.

The robot's energy is modified according to the values

given in each cell.

The robot stops if there are no more cells with rewards or if its energy runs out.

Actions

Design the robot's behavior according to the following rules, considering the updated information:

The robot moves randomly in the room.

The robot selects the cell with the highest reward and traces a route to collect it

$.$

The robot searches for the nearest cell with a reward and traces a route to collect it

Add some graphs to visualize The answer

$.$

$$Parameters in

$.$

json:

$\{$

$"$

money

$"$

:

$1$

$,$

$"$position$"$:

$[$

$0$

$,$

$4$

$]$

$,$

$"$step

$\_$

cost

$"$

:

$1$

$,$

$"$state$"$:

$[$

$[$

$5$

$,$

$0$

$,$

$0$

$,$

$3$

$,$

$0$

$,$

$2$

$,$

$0$

$,$

$0$

$,$

$1$

$]$

$]$

$\}$

ParameterViewer:

import json, numpy as np

from matplotlib import pyplot as plt

import seaborn as sns

# Reading information from files

f $=$ open$(\text{'}$params$3.$json'$)$

parameters $=$ json.load$($f$)$

state, money, step$\_$cost$,$ position $=[$parameters$[$key$]$ for key in $[\text{'}$state$\text{'},$ 'money', 'step$\_$cost$\text{'},$ 'position'$]]$

h$,$ w $=$ np$.$array$($state$).$shape

plt$.$rcParams$[\text{'}$figure$.$figsize'$]=[$w$//2,$ h$//2]$

print$(\text{'}$money: $\{\}\text{'}.$format$($money$))$

print$(\text{'}$step cost: $\{\}\text{'}.$format$($step$\_$cost$))$

print$(\text{'}$position: $\{\}\text{'}.$format$($position$))$

ax $=$ sns$.$heatmap$($state$,$ linewidth$=0\mathrm{.}5,$ linecolor$=$'black', annot$=$True, cmap$=$'RdBu', cbar$=$False, vmin$=-7,$ vmax$=7)$

plt$.$xticks$([])$

plt$.$yticks$([])$

plt$.$show$()$