Psi Chi R - June 2024

Source: Psi Chi R

Load packages and import data

knitr::opts_chunk$set(echo = TRUE,warning=F)
#library(tidyverse)

library(dplyr)
library(purrr)
library(ggplot2)

dataset=read.csv('https://osf.io/download/c3bfs/')

Check the data

head(dataset)

dexnum	name	generation	type1	type2	species	height	weight	ability1	ability2	hidden_ability	hp	attack	defense	sp_atk	sp_def	speed	total	ev_yield	catch_rate	base_friendship	base_exp	growth_rate	egg_group1	egg_group2	percent_male	percent_female	egg_cycles	special_group
1	Bulbasaur	1	Grass	Poison	Seed Pokémon	0.7	6.9	Overgrow	Chlorophyll		45	49	49	65	65	45	318	1 Sp. Atk	45	50	64	Medium Slow	Grass	Monster	87.5	12.5	20	Ordinary
2	Ivysaur	1	Grass	Poison	Seed Pokémon	1.0	13.0	Overgrow	Chlorophyll		60	62	63	80	80	60	405	1 Sp. Atk, 1 Sp. Def	45	50	142	Medium Slow	Grass	Monster	87.5	12.5	20	Ordinary
3	Venusaur	1	Grass	Poison	Seed Pokémon	2.0	100.0	Overgrow	Chlorophyll		80	82	83	100	100	80	525	2 Sp. Atk, 1 Sp. Def	45	50	236	Medium Slow	Grass	Monster	87.5	12.5	20	Ordinary
4	Charmander	1	Fire		Lizard Pokémon	0.6	8.5	Blaze	Solar Power		39	52	43	60	50	65	309	1 Speed	45	50	62	Medium Slow	Dragon	Monster	87.5	12.5	20	Ordinary
5	Charmeleon	1	Fire		Flame Pokémon	1.1	19.0	Blaze	Solar Power		58	64	58	80	65	80	405	1 Sp. Atk, 1 Speed	45	50	142	Medium Slow	Dragon	Monster	87.5	12.5	20	Ordinary
6	Charizard	1	Fire	Flying	Flame Pokémon	1.7	90.5	Blaze	Solar Power		78	84	78	109	85	100	534	3 Sp. Atk	45	50	267	Medium Slow	Dragon	Monster	87.5	12.5	20	Ordinary

skimr::skim(dataset)

Data summary

Name	dataset
Number of rows	1025
Number of columns	29
_______________________
Column type frequency:
character	15
numeric	14
________________________
Group variables	None

Variable type: character

skim_variable	n_missing	complete_rate	min	max	empty	n_unique	whitespace
name	0	1	3	17	0	1025	0
type1	0	1	3	8	0	18	0
type2	0	1	0	8	499	19	0
species	0	1	1	21	0	715	0
ability1	0	1	4	16	0	223	0
ability2	0	1	0	16	167	170	0
hidden_ability	0	1	0	13	495	126	0
ev_yield	0	1	4	30	0	48	0
base_friendship	0	1	1	3	0	9	0
base_exp	0	1	1	3	0	181	0
growth_rate	0	1	4	11	0	6	0
egg_group1	0	1	3	12	0	26	0
egg_group2	0	1	0	10	746	12	0
egg_cycles	0	1	1	3	0	12	0
special_group	0	1	6	15	0	8	0

Variable type: numeric

skim_variable	n_missing	complete_rate	mean	sd	p0	p25	p50	p75	p100	hist
dexnum	0	1.00	513.00	296.04	1.0	257.0	513	769.0	1025.0	▇▇▇▇▇
generation	0	1.00	4.74	2.63	1.0	3.0	5	7.0	9.0	▇▇▅▅▇
height	0	1.00	1.21	1.25	0.1	0.5	1	1.5	20.0	▇▁▁▁▁
weight	0	1.00	66.98	121.28	0.1	8.5	28	70.0	999.9	▇▁▁▁▁
hp	0	1.00	70.18	26.63	1.0	50.0	68	85.0	255.0	▃▇▁▁▁
attack	0	1.00	77.52	29.78	5.0	55.0	75	100.0	181.0	▂▇▆▃▁
defense	0	1.00	72.51	29.29	5.0	50.0	70	90.0	230.0	▃▇▂▁▁
sp_atk	0	1.00	70.08	29.66	10.0	47.0	65	90.0	173.0	▃▇▅▂▁
sp_def	0	1.00	70.21	26.64	20.0	50.0	67	86.0	230.0	▇▇▂▁▁
speed	0	1.00	67.19	28.72	5.0	45.0	65	88.0	200.0	▃▇▅▁▁
total	0	1.00	427.69	112.77	175.0	325.0	450	508.0	720.0	▂▆▇▆▁
catch_rate	0	1.00	94.93	76.11	3.0	45.0	60	140.0	255.0	▇▃▂▂▂
percent_male	155	0.85	54.93	20.31	0.0	50.0	50	50.0	100.0	▁▁▇▁▂
percent_female	155	0.85	45.07	20.31	0.0	50.0	50	50.0	100.0	▂▁▇▁▁

# sum(is.na(dataset))
# 
# sum(is.null(dataset))
# 
# sum(!complete.cases(dataset))

Level 1-

Let’s get familiar with our data. Create a table that shows the frequency of each type of Pokémon in the dataset (variable name: type1).

tab1 = dataset %>% 
  filter(complete.cases(type1)) %>% 
  group_by(type1) %>% 
  count() %>% 
  arrange(desc(n))

tab1

type1	n
Water	134
Normal	118
Grass	103
Bug	83
Fire	66
Psychic	60
Electric	59
Rock	58
Dark	45
Poison	42
Fighting	40
Ground	40
Dragon	37
Steel	36
Ghost	35
Ice	31
Fairy	29
Flying	9

Next, let’s look at the representation of each generation of Pokémon in the dataset. Create a table that shows the number of Pokémon in the dataset associated with each generation (variable name: generation)

tab2 = dataset %>% 
  filter(complete.cases(generation)) %>% 
  group_by(generation) %>% 
  count() %>% 
  arrange(desc(n))

tab2

generation	n
5	156
1	151
3	135
9	120
4	107
2	100
8	96
7	88
6	72

Level 2-

Provide the mean, standard deviation, median, and range of values of attack stats (variable name: attack) for ‘Water’ type Pokémon (variable name: type1).

water_attack = dataset %>% 
  filter(type1 == 'Water',
         complete.cases(type1)) %>% 
  select(attack)

#create function for desc stats
desc_stat = function(x){
  c(mean = mean(x,na.rm = T),
  standard_dev = sd(x,na.rm = T),
  median = median(x,na.rm = T),
  range = range(x,na.rm = T))
}

#mean: 72.34
#std: 25.26
#median: 70
#range: 10 to 130
map(water_attack,desc_stat)

## $attack
##         mean standard_dev       median       range1       range2 
##     72.34328     25.25959     70.00000     10.00000    130.00000

-Provide the mean, standard deviation, median, and range of values of speed stats (variable name: speed) for ‘Dragon’ type Pokémon (variable name: type1).

dragon_speed = dataset %>% 
  filter(type1 == 'Dragon',
         complete.cases(type1)) %>% 
  select(speed)


#mean: 80.19
#std: 24.62
#median: 82
#range: 40 to 142
map(dragon_speed,desc_stat)

## $speed
##         mean standard_dev       median       range1       range2 
##     80.18919     24.62365     82.00000     40.00000    142.00000

Level 3-

Are there any outliers for the ‘total’ stat? What is the value(s), and which Pokémon(s) is/are associated with it?

library(ggplot2)

boxplot(dataset$total)

outlier_lower = dataset %>% 
  select(name,total) %>% 
  arrange(total)

outlier_upper = dataset %>% 
  select(name,total) %>% 
  arrange(desc(total))

head(outlier_lower)

name	total
Wishiwashi	175
Sunkern	180
Blipbug	180
Snom	185
Azurill	190
Kricketot	194

head(outlier_upper)

name	total
Arceus	720
Eternatus	690
Mewtwo	680
Lugia	680
Ho-oh	680
Rayquaza	680

#Wishwashi with a total of 175 is the lower outlier and Arceus with a total of 720 is the upper outlier.

##Level 4-

Is there a correlation between a Pokémon’s ‘total’ stat and their base level of experience? (variable: base_exp)

dataset2=dataset %>% 
  filter(complete.cases(base_exp),
         complete.cases(total)) %>% 
  mutate(base_exp2=as.numeric(base_exp)) %>% 
  filter(!is.na(base_exp2),
         !is.na(total)) %>% 
  select(base_exp,base_exp2,total,everything())

#violates normality
shapiro.test(dataset2$base_exp2)

## 
##  Shapiro-Wilk normality test
## 
## data:  dataset2$base_exp2
## W = 0.92187, p-value < 2.2e-16

hist(dataset2$base_exp2)

# p-value < 2.2e-16, sample estimates: tau 0.9132707 
cor.test(dataset2$total,dataset2$base_exp2,method = 'kendall')

## 
##  Kendall's rank correlation tau
## 
## data:  dataset2$total and dataset2$base_exp2
## z = 42.937, p-value < 2.2e-16
## alternative hypothesis: true tau is not equal to 0
## sample estimates:
##       tau 
## 0.9132707

-Create a visualization that shows the average ‘total’ stat based on the primary type of the Pokémon (type1)

dataset3 = dataset2 %>% 
  group_by('Pokemon Primary Type' =type1) %>% 
  summarise('Average of Total Stat'=round(mean(total),2)) %>% 
  arrange(desc(`Average of Total Stat`))

dataset3

Pokemon Primary Type	Average of Total Stat
Dragon	490.16
Steel	468.03
Dark	454.73
Psychic	446.72
Fire	443.98
Fighting	441.55
Rock	438.54
Ice	436.39
Flying	436.11
Fairy	436.07
Ground	434.58
Electric	433.66
Ghost	431.17
Water	417.58
Poison	411.61
Grass	407.94
Normal	399.41
Bug	374.64

dataset3 %>% 
  ggplot(aes(y=`Pokemon Primary Type`,x=`Average of Total Stat`,fill=`Average of Total Stat`))+
  geom_col()+
  labs(x='Total Stat Average',y='Primary Type',
       title='A Bar Chart Brawl of Pokemon Power',
       subtitle='Examining Primary Types and their Average Total Stats',
       fill=' ')+
  theme_bw()+
  theme(plot.title = element_text(hjust=.5),
        plot.subtitle = element_text(hjust = .5))