Tabelle di contingenza

Inferenza per dati categorici in R

Andrew Bray

Assistant Professor of Statistics at Reed College

Politica e spesa militare

gss2016 %>%
  select(party, natarms) %>%
  glimpse()

Osservazioni: 150
Variabili: 2
$ party   <fct> Ind, Ind, Dem, Ind, Ind, Ind, Ind, Dem, Dem, Ind,...
$ natarms <fct> TOO LITTLE, TOO MUCH, TOO MUCH,...

Politica e spesa militare

ggplot(gss2016, aes(x = party, fill = natarms)) +
  geom_bar(position = "fill")

Politica e spesa militare

ggplot(gss2016, aes(x = party, fill = natarms)) +
  geom_bar()

library(broom)

Tabelle e dati tidy

tab <- gss2016 %>%
  select(natarms, party) %>%
  table()
tab

             party
natarms          Dem  Ind  Rep  Oth
  TOO LITTLE     17    20   24    0
  ABOUT RIGHT    14    28    8    1
  TOO MUCH       12    24    2    0

Tabelle e dati tidy

tab <- gss2016 %>%
  select(natarms, party) %>%
  table()
tab

             party
natarms          Dem  Ind  Rep  Oth
  TOO LITTLE     17    20   24    0
  ABOUT RIGHT    14    28    8    1
  TOO MUCH       12    24    2    0

tab %>%
  tidy()

# A tibble: 12 x 3
   natarms     party     n
   <chr>       <chr> <int>
 1 TOO LITTLE  Dem      17
 2 ABOUT RIGHT Dem      14
 3 TOO MUCH    Dem      12
 4 TOO LITTLE  Ind      20
 5 ABOUT RIGHT Ind      28
 6 TOO MUCH    Ind      24
 7 TOO LITTLE  Rep      24
 8 ABOUT RIGHT Rep       8
 9 TOO MUCH    Rep       2
10 TOO LITTLE  Oth       0
11 ABOUT RIGHT Oth       1
12 TOO MUCH    Oth       0

Tabelle e dati tidy

tab <- gss2016 %>%
  select(natarms, party) %>%
  table()
tab

             party
natarms          Dem  Ind  Rep  Oth
  TOO LITTLE     17    20   24    0
  ABOUT RIGHT    14    28    8    1
  TOO MUCH       12    24    2    0

tab %>%
  tidy() %>%
  uncount(n)

# A tibble: 150 x 2
   natarms    party
   <chr>      <chr>
 1 TOO LITTLE Dem  
 2 TOO LITTLE Dem  
 3 TOO LITTLE Dem  
 4 TOO LITTLE Dem  
 5 TOO LITTLE Dem  
 6 TOO LITTLE Dem  
 7 TOO LITTLE Dem  
 8 TOO LITTLE Dem  
 9 TOO LITTLE Dem  
10 TOO LITTLE Dem  
# … with 140 more rows

Ayo berlatih!

Inferenza per dati categorici in R