Week 1: Mushrooms

First, we load the data from the online source and save it as a data frame. We then subset it to four variables, name those columns, and replace the abbreviations with full descriptional values.

src <- 'https://archive.ics.uci.edu/ml/machine-learning-databases/mushroom/agaricus-lepiota.data'
mush <- read.csv(url(src), header=FALSE, stringsAsFactors = FALSE)
mushroom <- as.data.frame(mush)
newMush <- mushroom[c(1,19, 22:23)] #subset the data, keeping four columns
names(newMush) <- c("classes", "ringNumber", "population", "habitat") 
#renaming the values based on the provided data dictionary
newMush$classes[newMush$classes=='p'] <- 'poisonous'
newMush$classes[newMush$classes=='e'] <- 'edible'
newMush$ringNumber[newMush$ringNumber=='n'] <- 'none'
newMush$ringNumber[newMush$ringNumber=='o'] <- 'one'
newMush$ringNumber[newMush$ringNumber=='t'] <- 'two'
newMush$population[newMush$population=='a'] <- 'abundant'
newMush$population[newMush$population=='c'] <- 'clustered'
newMush$population[newMush$population=='n'] <- 'numerous'
newMush$population[newMush$population=='s'] <- 'scattered'
newMush$population[newMush$population=='v'] <- 'several'
newMush$population[newMush$population=='y'] <- 'solitary'
newMush$habitat[newMush$habitat=='g'] <- 'grasses'
newMush$habitat[newMush$habitat=='l'] <- 'leaves'
newMush$habitat[newMush$habitat=='m'] <- 'meadows'
newMush$habitat[newMush$habitat=='p'] <- 'paths'
newMush$habitat[newMush$habitat=='u'] <- 'urban'
newMush$habitat[newMush$habitat=='w'] <- 'waste'
newMush$habitat[newMush$habitat=='d'] <- 'woods'
head(newMush)

##     classes ringNumber population habitat
## 1 poisonous        one  scattered   urban
## 2    edible        one   numerous grasses
## 3    edible        one   numerous meadows
## 4 poisonous        one  scattered   urban
## 5    edible        one   abundant grasses
## 6    edible        one   numerous grasses

Looking at contingency tables, it seems that the population type with the largest number of poisonous mushrooms is ‘several’, and the habitat with the largest number of poisonous mushrooms is ‘woods’.

table(newMush$classes, newMush$population)

##            
##             abundant clustered numerous scattered several solitary
##   edible         384       288      400       880    1192     1064
##   poisonous        0        52        0       368    2848      648

table(newMush$classes, newMush$habitat)

##            
##             grasses leaves meadows paths urban waste woods
##   edible       1408    240     256   136    96   192  1880
##   poisonous     740    592      36  1008   272     0  1268