class: center, middle, inverse, title-slide # Trabalho Diagrama de dispersão e Matriz de correlação ## ⚔<br/> com o xaringan ### Leir Gustavo ### UNIRIO ### 23/04/2021 --- --- class: center, middle # Análise de duas variáveis quantitativas ### Diagrama de dispersão ### Coeficiente de correlação --- # Passo 1 - Carregar a base de dados ```r library(readr) poke <- read_csv("C:/Users/favor/Base_de_dados-master/pokRdex_mod.csv") ``` ``` ## ## -- Column specification -------------------------------------------------------- ## cols( ## .default = col_character(), ## id = col_double(), ## species_id = col_double(), ## height = col_double(), ## weight = col_double(), ## base_experience = col_double(), ## attack = col_double(), ## defense = col_double(), ## hp = col_double(), ## special_attack = col_double(), ## special_defense = col_double(), ## speed = col_double(), ## generation_id = col_double(), ## evolves_from_species_id = col_double(), ## evolution_chain_id = col_double(), ## shape_id = col_double() ## ) ## i Use `spec()` for the full column specifications. ``` ```r library(dplyr) ``` ``` ## ## Attaching package: 'dplyr' ``` ``` ## The following objects are masked from 'package:stats': ## ## filter, lag ``` ``` ## The following objects are masked from 'package:base': ## ## intersect, setdiff, setequal, union ``` --- # Passo 2 - Carregar a biblioteca DPLYR ```r library(dplyr) ``` --- # Passo 4 - Resumo das variáveis ```r poke %>% summarise(media_altura=mean(height), media_peso=mean(weight), media_ataque=mean(attack), media_defesa=mean(defense), dp_ataque=sd(attack), dp_defesa=sd(defense), dp_altura=sd(height), dp_peso=sd(weight)) ``` ``` ## # A tibble: 1 x 8 ## media_altura media_peso media_ataque media_defesa dp_ataque dp_defesa ## <dbl> <dbl> <dbl> <dbl> <dbl> <dbl> ## 1 17.7 620. 78.7 73.7 32.0 31.0 ## # ... with 2 more variables: dp_altura <dbl>, dp_peso <dbl> ``` --- # Passo 5 - Diagrama de dispersão <!-- --><!-- --><!-- --> --- # Correlação entre as váriáveis ```r cor(poke$height,poke$weight) ``` ``` ## [1] 0.403526 ``` ```r cor(poke$height,poke$attack) ``` ``` ## [1] 0.2539995 ``` ```r cor(poke$height,poke$defense) ``` ``` ## [1] 0.246898 ``` --- # Matriz de correlação entre as variáveis ```r poke_quanti <- poke %>% select(height,weight,attack,defense) cor(poke_quanti) ``` ``` ## height weight attack defense ## height 1.0000000 0.4035260 0.2539995 0.2468980 ## weight 0.4035260 1.0000000 0.4487842 0.4885063 ## attack 0.2539995 0.4487842 1.0000000 0.4355845 ## defense 0.2468980 0.4885063 0.4355845 1.0000000 ``` ```r library(corrplot) ``` ``` ## Warning: package 'corrplot' was built under R version 4.0.5 ``` ``` ## corrplot 0.84 loaded ``` ```r MCorr <- cor(poke_quanti) corrplot(MCorr) ``` <!-- --> --- # Conclusão: ## Como apresentado na matriz e no diagrama de dispersão, as correlações entre peso, altura, defesa e ataque são fracas, tendo no máximo uma correlação de 0.4, o que a colocaria como uma correlação fraca.