(GIF: https://git.io/vAC88, Data: http://archive.ics.uci.edu/ml/machine-learning-databases/mushroom/agaricus-lepiota.data) Label columns as factors using a bit of Vim/Spacemacs foo

mushroom-label.r

# Assumes an existing data frame called `raw`
# that is directly imported from the URL and
# has not converted the "?" values to NA's

colnames(raw) <- c("Type", "CapShape", "CapSurface", "CapColor", "Bruises",
                   "Odor", "GillAttachment", "GillSpacing", "GillSize",
                   "GillColor", "StalkShape", "StalkRoot",
                   "StalkSurfaceAboveRing", "StalkSurfaceBelowRing",
                   "StalkColorAboveRing", "StalkColorBelowRing", "VeilType",
                   "VeilColor", "RingNumber", "RingType", "SporePrintColor",
                   "Population", "Habitat")

raw$Type <- factor(raw$Type,
                   levels = c("e", "p"),
                   labels = c("edible", "poisonous"))

raw$CapShape <- factor(raw$CapShape,
                       levels = c("b", "c", "x", "f", "k", "s"),
                       labels = c("bell", "conical", "convex", "flat",
                                  "knobbed", "sunken"))

raw$CapSurface <- factor(raw$CapSurface,
                         levels = c("f", "g", "y", "s"),
                         labels = c("fibrous", "grooves", "scaly", "smooth"))

raw$CapColor <- factor(raw$CapColor,
                       levels = c("n", "b", "c", "g", "r", "p", "u", "e", "w",
                                  "y"),
                       labels = c("brown", "buff", "cinnamon", "gray", "green",
                                  "pink", "purple", "red", "white", "yellow"))

raw$Bruises <- factor(raw$Bruises,
                      levels = c("t", "f"),
                      labels = c("bruises", "no"))

raw$Odor <- factor(raw$Odor,
                   levels = c("a", "l", "c", "y", "f", "m", "n", "p", "s"),
                   labels = c("almond", "anise", "creosote", "fishy", "foul",
                              "musty", "none", "pungent", "spicy"))

raw$GillAttachment <- factor(raw$GillAttachment,
                             levels = c("a", "d", "f", "n"),
                             labels = c("attached", "descending", "free",
                                        "notched"))

raw$GillSpacing <- factor(raw$GillSpacing,
                          levels = c("c", "w", "d"),
                          labels = c("close", "crowded", "distant"))

raw$GillSize <- factor(raw$GillSize,
                       levels = c("b", "n"),
                       labels = c("broad", "narrow"))

raw$GillColor <- factor(raw$GillColor,
                        levels = c("k", "n", "b", "h", "g", "r", "o", "p", "u",
                                   "e", "w", "y"),
                        labels = c("black", "brown", "buff", "chocolate", "gray",
                                   "green", "orange", "pink", "purple", "red",
                                   "white", "yellow"))

raw$StalkShape <- factor(raw$StalkShape,
                         levels = c("e", "t"),
                         labels = c("enlarging", "tapering"))

raw$StalkRoot <- factor(raw$StalkRoot,
                        levels = c("b", "c", "u", "e", "z", "r", "?"),
                        labels = c("bulbous", "club", "cup", "equal",
                                   "rhizomorphs", "rooted", "missing"))

raw$StalkSurfaceAboveRing <- factor(raw$StalkSurfaceAboveRing,
                                    levels = c("f", "y", "k", "s"),
                                    labels = c("fibrous", "scaly", "silky",
                                               "smooth"))

raw$StalkSurfaceBelowRing <- factor(raw$StalkSurfaceBelowRing,
                                    levels = c("f", "y", "k", "s"),
                                    labels = c("fibrous", "scaly", "silky",
                                               "smooth"))

raw$StalkColorAboveRing <- factor(raw$StalkColorAboveRing,
                                  levels = c("n", "b", "c", "g", "o", "p", "e",
                                             "w", "y"),
                                  labels = c("brown", "buff", "cinnamon", "gray",
                                             "orange", "pink", "red", "white",
                                             "yellow"))

raw$StalkColorBelowRing <- factor(raw$StalkColorBelowRing,
                                  levels = c("n", "b", "c", "g", "o", "p", "e",
                                             "w", "y"),
                                  labels = c("brown", "buff", "cinnamon",
                                             "gray", "orange", "pink", "red",
                                             "white", "yellow"))

raw$VeilType <- factor(raw$VeilType,
                       levels = c("p", "u"),
                       labels = c("partial", "universal"))

raw$VeilColor <- factor(raw$VeilColor,
                        levels = c("n", "o", "w", "y"),
                        labels = c("brown", "orange", "white", "yellow"))

raw$RingNumber <- factor(raw$RingNumber,
                         levels = c("n", "o", "t"),
                         labels = c("none", "one", "two"))

raw$RingType <- factor(raw$RingType,
                       levels = c("c", "e", "f", "l", "n", "p", "s", "z"),
                       labels = c("cobwebby", "evanescent", "flaring",
                                  "large", "none", "pendant", "sheathing",
                                  "zone"))

raw$SporePrintColor <- factor(raw$SporePrintColor,
                              levels = c("k", "n", "b", "h", "r", "o", "u",
                                         "w", "y"),
                              labels = c("black", "brown", "buff", "chocolate",
                                         "green", "orange", "purple", "white",
                                         "yellow"))

raw$Population <- factor(raw$Population,
                         levels = c("a", "c", "n", "s", "v", "y"),
                         labels = c("abundant", "clustered", "numerous",
                                    "scattered", "several", "solitary"))

raw$Habitat <- factor(raw$Habitat,
                      levels = c("g", "l", "m", "p", "u", "w", "d"),
                      labels = c("grasses", "leaves", "meadows", "paths",
                                 "urban", "waste", "woods"))