_CompareMeans.Rmd

# My R Codes For Data Analysis 
sub# In this repository I am going to collect `R codes` for data analysis. Codes are from various resources and I try to give original link as much as possible. 
author:  [Serdar Balcı, MD, Pathologist](https://www.serdarbalci.com/) 
date: '`{r #  format(Sys.Date())`'



### Compare Means

```
{r eval=FALSE, include=FALSE, echo=TRUE}
t.test(scabies$age[scabies$gender== male ],scabies$age[scabies$gender== female ])

```

```
{r eval=FALSE, include=FALSE, echo=TRUE}
test <- t.test(scabies$age[scabies$gender== male ],scabies$age[scabies$gender== female ])
psycho::analyze(test)
```


# infer

Randomization Examples using nycflights13 flights data

https://cran.r-project.org/web/packages/infer/vignettes/flights_examples.html


```
{r eval=FALSE, include=FALSE, echo=TRUE}
library(nycflights13)
library(dplyr)
library(ggplot2)
library(stringr)
library(infer)
set.seed(2017)
```


```
{r eval=FALSE, include=FALSE, echo=TRUE}
fli_small <- flights %>% 
  na.omit() %>%
  sample_n(size = 500) %>% 
  mutate(season = case_when(
    month %in% c(10:12, 1:3) ~  winter ,
    month %in% c(4:9) ~  summer 
  )) %>% 
  mutate(day_hour = case_when(
    between(hour, 1, 12) ~  morning ,
    between(hour, 13, 24) ~  not morning 
  )) %>% 
  select(arr_delay, dep_delay, season, 
         day_hour, origin, carrier)
fli_small
```


Hypothesis tests
One numerical variable (mean)


```
{r eval=FALSE, include=FALSE, echo=TRUE}
x_bar <- fli_small %>%
  summarize(mean(dep_delay)) %>%
  pull()
x_bar
```


```
{r eval=FALSE, include=FALSE, echo=TRUE}
null_distn <- fli_small %>%
  specify(response = dep_delay) %>%
  hypothesize(null =  point , mu = 10) %>%
  generate(reps = 1000, type =  bootstrap ) %>%
  calculate(stat =  mean )
null_distn
```


```
{r eval=FALSE, include=FALSE, echo=TRUE}
ggplot(data = null_distn, mapping = aes(x = stat)) +
  geom_density() +
  geom_vline(xintercept = x_bar, color =  red )
```


```
{r eval=FALSE, include=FALSE, echo=TRUE}
null_distn %>%
  summarize(p_value = mean(stat >= x_bar) * 2)
```


One numerical variable (median)

```
{r eval=FALSE, include=FALSE, echo=TRUE}
x_tilde <- fli_small %>%
  summarize(median(dep_delay)) %>%
  pull()
x_tilde
```


```
{r eval=FALSE, include=FALSE, echo=TRUE}
null_distn <- fli_small %>%
  specify(response = dep_delay) %>%
  hypothesize(null =  point , med = -1) %>% 
  generate(reps = 1000, type =  bootstrap ) %>% 
  calculate(stat =  median )
null_distn
```


```
{r eval=FALSE, include=FALSE, echo=TRUE}
ggplot(null_distn, aes(x = stat)) +
  geom_bar() +
  geom_vline(xintercept = x_tilde, color =  red )
```


```
{r eval=FALSE, include=FALSE, echo=TRUE}
null_distn %>%
  summarize(p_value = mean(stat <= x_tilde) * 2)
```



One categorical (one proportion)


```
{r eval=FALSE, include=FALSE, echo=TRUE}
p_hat <- fli_small %>%
  summarize(mean(day_hour ==  morning )) %>%
  pull()
p_hat
```





```
null_distn <- fli_small %>%
  specify(response = day_hour, success =  morning ) %>%
  hypothesize(null =  point , p = .5) %>%
  generate(reps = 1000, type =  simulate ) %>%
  calculate(stat =  prop )
ggplot(null_distn, aes(x = stat)) +
  geom_bar() +
  geom_vline(xintercept = p_hat, color =  red )

null_distn %>%
  summarize(p_value = mean(stat <= p_hat) * 2)
p_value
0.132
Logical variables will be coerced to factors:

null_distn <- fli_small %>%
  mutate(day_hour_logical = (day_hour ==  morning )) %>%
  specify(response = day_hour_logical, success =  TRUE ) %>%
  hypothesize(null =  point , p = .5) %>%
  generate(reps = 1000, type =  simulate ) %>%
  calculate(stat =  prop )
Two categorical (2 level) variables
d_hat <- fli_small %>%
  group_by(season) %>%
  summarize(prop = mean(day_hour ==  morning )) %>%
  summarize(diff(prop)) %>%
  pull()
null_distn <- fli_small %>%
  specify(day_hour ~ season, success =  morning ) %>%
  hypothesize(null =  independence ) %>% 
  generate(reps = 1000, type =  permute ) %>% 
  calculate(stat =  diff in props , order = c( winter ,  summer ))
ggplot(null_distn, aes(x = stat)) +
  geom_density() +
  geom_vline(xintercept = d_hat, color =  red )

null_distn %>%
  summarize(p_value = mean(stat <= d_hat) * 2) %>%
  pull()
## [1] 0.758
One categorical (>2 level) - GoF
Chisq_hat <- fli_small %>%
  specify(response = origin) %>%
  hypothesize(null =  point , 
              p = c( EWR  = .33,  JFK  = .33,  LGA  = .34)) %>% 
  calculate(stat =  Chisq )
null_distn <- fli_small %>%
  specify(response = origin) %>%
  hypothesize(null =  point , 
              p = c( EWR  = .33,  JFK  = .33,  LGA  = .34)) %>% 
  generate(reps = 1000, type =  simulate ) %>% 
  calculate(stat =  Chisq )
ggplot(null_distn, aes(x = stat)) +
  geom_density() +
  geom_vline(xintercept = pull(Chisq_hat), color =  red )

null_distn %>%
  summarize(p_value = mean(stat >= pull(Chisq_hat))) %>%
  pull()
## [1] 0.002
Two categorical (>2 level) variables
Chisq_hat <- fli_small %>%
  chisq_stat(formula = day_hour ~ origin)
null_distn <- fli_small %>%
  specify(day_hour ~ origin, success =  morning ) %>%
  hypothesize(null =  independence ) %>% 
  generate(reps = 1000, type =  permute ) %>% 
  calculate(stat =  Chisq )
ggplot(null_distn, aes(x = stat)) +
  geom_density() +
  geom_vline(xintercept = pull(Chisq_hat), color =  red )

null_distn %>%
  summarize(p_value = mean(stat >= pull(Chisq_hat))) %>%
  pull()
## [1] 0.017
One numerical variable, one categorical (2 levels) (diff in means)
d_hat <- fli_small %>% 
  group_by(season) %>% 
  summarize(mean_stat = mean(dep_delay)) %>% 
  # Since summer - winter
  summarize(-diff(mean_stat)) %>% 
  pull()
null_distn <- fli_small %>%
  specify(dep_delay ~ season) %>% # alt: response = dep_delay, 
  # explanatory = season
  hypothesize(null =  independence ) %>%
  generate(reps = 1000, type =  permute ) %>%
  calculate(stat =  diff in means , order = c( summer ,  winter ))
ggplot(null_distn, aes(x = stat)) +
  geom_density() +
  geom_vline(xintercept = d_hat, color =  red )

null_distn %>%
  summarize(p_value = mean(stat <= d_hat) * 2) %>%
  pull()
## [1] 1.574
One numerical variable, one categorical (2 levels) (diff in medians)
d_hat <- fli_small %>% 
  group_by(season) %>% 
  summarize(median_stat = median(dep_delay)) %>% 
  # Since summer - winter
  summarize(-diff(median_stat)) %>% 
  pull()
null_distn <- fli_small %>%
  specify(dep_delay ~ season) %>% # alt: response = dep_delay, 
  # explanatory = season
  hypothesize(null =  independence ) %>%
  generate(reps = 1000, type =  permute ) %>%
  calculate(stat =  diff in medians , order = c( summer ,  winter ))
ggplot(null_distn, aes(x = stat)) +
  geom_bar() +
  geom_vline(xintercept = d_hat, color =  red )

null_distn %>%
  summarize(p_value = mean(stat >= d_hat) * 2) %>%
  pull()
## [1] 0.068
One numerical, one categorical (>2 levels) - ANOVA
F_hat <- anova(
               aov(formula = arr_delay ~ origin, data = fli_small)
               )$`F value`[1]
null_distn <- fli_small %>%
   specify(arr_delay ~ origin) %>% # alt: response = arr_delay, 
   # explanatory = origin
   hypothesize(null =  independence ) %>%
   generate(reps = 1000, type =  permute ) %>%
   calculate(stat =  F )
ggplot(null_distn, aes(x = stat)) +
  geom_density() +
  geom_vline(xintercept = F_hat, color =  red )  

null_distn %>% 
  summarize(p_value = mean(stat >= F_hat)) %>%
  pull()
## [1] 0.351
Two numerical vars - SLR
slope_hat <- lm(arr_delay ~ dep_delay, data = fli_small) %>% 
  broom::tidy() %>% 
  filter(term ==  dep_delay ) %>% 
  pull(estimate)
null_distn <- fli_small %>%
   specify(arr_delay ~ dep_delay) %>% 
   hypothesize(null =  independence ) %>%
   generate(reps = 1000, type =  permute ) %>%
   calculate(stat =  slope )
ggplot(null_distn, aes(x = stat)) +
  geom_density() +
  geom_vline(xintercept = slope_hat, color =  red )  

null_distn %>% 
  summarize(p_value = mean(stat >= slope_hat) * 2) %>%
  pull()
## [1] 0
Confidence intervals
One numerical (one mean)
x_bar <- fli_small %>%
   summarize(mean(arr_delay)) %>%
   pull()
boot <- fli_small %>%
   specify(response = arr_delay) %>%
   generate(reps = 1000, type =  bootstrap ) %>%
   calculate(stat =  mean ) %>%
   pull()
c(lower = x_bar - 2 * sd(boot),
  upper = x_bar + 2 * sd(boot))
##    lower    upper 
## 1.122209 8.021791
One categorical (one proportion)
p_hat <- fli_small %>%
 summarize(mean(day_hour ==  morning )) %>%
 pull()
boot <- fli_small %>%
 specify(response = day_hour, success =  morning ) %>%
 generate(reps = 1000, type =  bootstrap ) %>%
 calculate(stat =  prop ) %>%
 pull()
c(lower = p_hat - 2 * sd(boot),
 upper = p_hat + 2 * sd(boot))
##     lower     upper 
## 0.4194756 0.5125244
One numerical variable, one categorical (2 levels) (diff in means)
d_hat <- fli_small %>% 
  group_by(season) %>% 
  summarize(mean_stat = mean(arr_delay)) %>% 
  # Since summer - winter
  summarize(-diff(mean_stat)) %>% 
  pull()
boot <- fli_small %>%
   specify(arr_delay ~ season) %>%
   generate(reps = 1000, type =  bootstrap ) %>%
   calculate(stat =  diff in means , order = c( summer ,  winter )) %>% 
   pull()
c(lower = d_hat - 2 * sd(boot), 
  upper = d_hat + 2 * sd(boot))
##     lower     upper 
## -7.704370  6.213971
Two categorical variables (diff in proportions)
d_hat <- fli_small %>%
  group_by(season) %>%
  summarize(prop = mean(day_hour ==  morning )) %>%
  # Since summer - winter
  summarize(-diff(prop)) %>%
  pull()
boot <- fli_small %>%
  specify(day_hour ~ season, success =  morning ) %>%
  generate(reps = 1000, type =  bootstrap ) %>% 
  calculate(stat =  diff in props , order = c( summer ,  winter )) %>%
  pull()
c(lower = d_hat - 2 * sd(boot), 
  upper = d_hat + 2 * sd(boot))
##       lower       upper 
## -0.07149487  0.11258550
Two numerical vars - SLR
slope_hat <- lm(arr_delay ~ dep_delay, data = fli_small) %>% 
  broom::tidy() %>% 
  filter(term ==  dep_delay ) %>% 
  pull(estimate)
boot <- fli_small %>%
   specify(arr_delay ~ dep_delay) %>% 
   generate(reps = 1000, type =  bootstrap ) %>%
   calculate(stat =  slope ) %>% 
   pull()
c(lower = slope_hat - 2 * sd(boot), 
  upper = slope_hat + 2 * sd(boot))   
##     lower     upper 
## 0.9657595 1.0681384






Examples using mtcars data


https://cran.r-project.org/web/packages/infer/vignettes/mtcars_examples.html


Examples using mtcars data
Chester Ismay and Andrew Bray
2018-01-05
Note: The type argument in generate() is automatically filled based on the entries for specify() and hypothesize(). It can be removed throughout the examples that follow. It is left in to reiterate the type of generation process being performed.

Data preparation
library(infer)
library(dplyr)
mtcars <- mtcars %>%
  mutate(cyl = factor(cyl),
         vs = factor(vs),
         am = factor(am),
         gear = factor(gear),
         carb = factor(carb))
# For reproducibility         
set.seed(2018)         
One numerical variable (mean)

mtcars %>%
  specify(response = mpg) %>% # formula alt: mpg ~ NULL
  hypothesize(null =  point , mu = 25) %>% 
  generate(reps = 100, type =  bootstrap ) %>% 
  calculate(stat =  mean )
## # A tibble: 100 x 2
##    replicate  stat
##        <int> <dbl>
##  1         1  26.6
##  2         2  25.1
##  3         3  25.2
##  4         4  24.7
##  5         5  24.6
##  6         6  25.8
##  7         7  24.7
##  8         8  25.6
##  9         9  25.0
## 10        10  25.1
## # ... with 90 more rows
One numerical variable (median)

mtcars %>%
  specify(response = mpg) %>% # formula alt: mpg ~ NULL
  hypothesize(null =  point , med = 26) %>% 
  generate(reps = 100, type =  bootstrap ) %>% 
  calculate(stat =  median )
## # A tibble: 100 x 2
##    replicate  stat
##        <int> <dbl>
##  1         1  28.2
##  2         2  27.2
##  3         3  26.2
##  4         4  26  
##  5         5  26.5
##  6         6  24.5
##  7         7  26  
##  8         8  28.2
##  9         9  28.2
## 10        10  23.2
## # ... with 90 more rows
One categorical (2 level) variable

mtcars %>%
  specify(response = am, success =  1 ) %>% # formula alt: am ~ NULL
  hypothesize(null =  point , p = .25) %>% 
  generate(reps = 100, type =  simulate ) %>% 
  calculate(stat =  prop )
## # A tibble: 100 x 2
##    replicate   stat
##    <fct>      <dbl>
##  1 1         0.375 
##  2 2         0.0625
##  3 3         0.125 
##  4 4         0.25  
##  5 5         0.188 
##  6 6         0.406 
##  7 7         0.219 
##  8 8         0.375 
##  9 9         0.344 
## 10 10        0.188 
## # ... with 90 more rows
Two categorical (2 level) variables

mtcars %>%
  specify(am ~ vs, success =  1 ) %>% # alt: response = am, explanatory = vs
  hypothesize(null =  independence ) %>%
  generate(reps = 100, type =  permute ) %>%
  calculate(stat =  diff in props , order = c( 0 ,  1 ))
## # A tibble: 100 x 2
##    replicate    stat
##        <int>   <dbl>
##  1         1 -0.421 
##  2         2 -0.167 
##  3         3 -0.421 
##  4         4 -0.0397
##  5         5  0.0873
##  6         6 -0.0397
##  7         7 -0.0397
##  8         8 -0.0397
##  9         9  0.0873
## 10        10 -0.167 
## # ... with 90 more rows
One categorical (>2 level) - GoF

mtcars %>%
  specify(cyl ~ NULL) %>% # alt: response = cyl
  hypothesize(null =  point , p = c( 4  = .5,  6  = .25,  8  = .25)) %>%
  generate(reps = 100, type =  simulate ) %>%
  calculate(stat =  Chisq )
## # A tibble: 100 x 2
##    replicate  stat
##    <fct>     <dbl>
##  1 1         6.75 
##  2 2         1.69 
##  3 3         3.19 
##  4 4         1.69 
##  5 5         6    
##  6 6         2.69 
##  7 7         4.75 
##  8 8         0.75 
##  9 9         0.688
## 10 10        3.69 
## # ... with 90 more rows
Two categorical (>2 level) variables

mtcars %>%
  specify(cyl ~ am) %>% # alt: response = cyl, explanatory = am
  hypothesize(null =  independence ) %>%
  generate(reps = 100, type =  permute ) %>%
  calculate(stat =  Chisq )
## # A tibble: 100 x 2
##    replicate  stat
##        <int> <dbl>
##  1         1 1.34 
##  2         2 1.63 
##  3         3 1.63 
##  4         4 2.63 
##  5         5 3.90 
##  6         6 1.74 
##  7         7 0.126
##  8         8 1.74 
##  9         9 1.34 
## 10        10 1.34 
## # ... with 90 more rows
One numerical variable one categorical (2 levels) (diff in means)

mtcars %>%
  specify(mpg ~ am) %>% # alt: response = mpg, explanatory = am
  hypothesize(null =  independence ) %>%
  generate(reps = 100, type =  permute ) %>%
  calculate(stat =  diff in means , order = c( 0 ,  1 ))
## # A tibble: 100 x 2
##    replicate   stat
##        <int>  <dbl>
##  1         1 -1.10 
##  2         2  0.217
##  3         3 -1.08 
##  4         4 -3.80 
##  5         5  3.08 
##  6         6  0.489
##  7         7  2.34 
##  8         8  4.10 
##  9         9 -1.86 
## 10        10 -0.210
## # ... with 90 more rows
One numerical variable one categorical (2 levels) (diff in medians)

mtcars %>%
  specify(mpg ~ am) %>% # alt: response = mpg, explanatory = am
  hypothesize(null =  independence ) %>%
  generate(reps = 100, type =  permute ) %>%
  calculate(stat =  diff in medians , order = c( 0 ,  1 ))
## # A tibble: 100 x 2
##    replicate  stat
##        <int> <dbl>
##  1         1  0.5 
##  2         2 -1.10
##  3         3  5.20
##  4         4  1.8 
##  5         5  0.5 
##  6         6  3.3 
##  7         7 -1.60
##  8         8 -2.3 
##  9         9  2.90
## 10        10 -0.5 
## # ... with 90 more rows
One numerical one categorical (>2 levels) - ANOVA

mtcars %>%
  specify(mpg ~ cyl) %>% # alt: response = mpg, explanatory = cyl
  hypothesize(null =  independence ) %>%
  generate(reps = 100, type =  permute ) %>%
  calculate(stat =  F )
## # A tibble: 100 x 2
##    replicate  stat
##        <int> <dbl>
##  1         1 1.43 
##  2         2 1.65 
##  3         3 0.318
##  4         4 0.393
##  5         5 1.05 
##  6         6 0.826
##  7         7 1.32 
##  8         8 0.833
##  9         9 0.144
## 10        10 0.365
## # ... with 90 more rows
Two numerical vars - SLR

mtcars %>%
  specify(mpg ~ hp) %>% # alt: response = mpg, explanatory = cyl
  hypothesize(null =  independence ) %>%
  generate(reps = 100, type =  permute ) %>%
  calculate(stat =  slope )
## # A tibble: 100 x 2
##    replicate     stat
##        <int>    <dbl>
##  1         1 -0.0151 
##  2         2  0.00224
##  3         3 -0.0120 
##  4         4  0.00292
##  5         5  0.0203 
##  6         6 -0.00730
##  7         7 -0.0246 
##  8         8  0.00555
##  9         9  0.0109 
## 10        10  0.0176 
## # ... with 90 more rows
One numerical variable (standard deviation)

Not currently implemented

mtcars %>%
  specify(response = mpg) %>% # formula alt: mpg ~ NULL
  hypothesize(null =  point , sigma = 5) %>% 
  generate(reps = 100, type =  bootstrap ) %>% 
  calculate(stat =  sd )
Confidence intervals
One numerical (one mean)

mtcars %>%
  specify(response = mpg) %>%
  generate(reps = 100, type =  bootstrap ) %>%
  calculate(stat =  mean )
## # A tibble: 100 x 2
##    replicate  stat
##        <int> <dbl>
##  1         1  19.6
##  2         2  21.8
##  3         3  18.7
##  4         4  19.2
##  5         5  21.6
##  6         6  19.9
##  7         7  20.7
##  8         8  19.3
##  9         9  21.2
## 10        10  21.3
## # ... with 90 more rows
One numerical (one median)

mtcars %>%
  specify(response = mpg) %>%
  generate(reps = 100, type =  bootstrap ) %>%
  calculate(stat =  median )
## # A tibble: 100 x 2
##    replicate  stat
##        <int> <dbl>
##  1         1  19.2
##  2         2  20.1
##  3         3  21  
##  4         4  17.8
##  5         5  20.1
##  6         6  19.2
##  7         7  18.4
##  8         8  19.2
##  9         9  19.2
## 10        10  18.0
## # ... with 90 more rows
One numerical (standard deviation)

mtcars %>%
  specify(response = mpg) %>%
  generate(reps = 100, type =  bootstrap ) %>%
  calculate(stat =  sd )
## # A tibble: 100 x 2
##    replicate  stat
##        <int> <dbl>
##  1         1  5.28
##  2         2  6.74
##  3         3  5.29
##  4         4  5.41
##  5         5  5.56
##  6         6  5.65
##  7         7  6.17
##  8         8  6.40
##  9         9  6.31
## 10        10  6.11
## # ... with 90 more rows
One categorical (one proportion)

mtcars %>%
  specify(response = am, success =  1 ) %>%
  generate(reps = 100, type =  bootstrap ) %>%
  calculate(stat =  prop )
## # A tibble: 100 x 2
##    replicate  stat
##        <int> <dbl>
##  1         1 0.375
##  2         2 0.406
##  3         3 0.406
##  4         4 0.312
##  5         5 0.312
##  6         6 0.469
##  7         7 0.438
##  8         8 0.281
##  9         9 0.438
## 10        10 0.5  
## # ... with 90 more rows
One numerical variable one categorical (2 levels) (diff in means)

mtcars %>%
  specify(mpg ~ am) %>%
  generate(reps = 100, type =  bootstrap ) %>%
  calculate(stat =  diff in means , order = c( 0 ,  1 ))
## # A tibble: 100 x 2
##    replicate   stat
##        <int>  <dbl>
##  1         1  -9.38
##  2         2  -5.11
##  3         3  -4.88
##  4         4  -5.39
##  5         5  -9.19
##  6         6  -7.20
##  7         7  -5.34
##  8         8  -3.20
##  9         9  -5.95
## 10        10 -11.0 
## # ... with 90 more rows
Two categorical variables (diff in proportions)

mtcars %>%
  specify(am ~ vs, success =  1 ) %>%
  generate(reps = 100, type =  bootstrap ) %>%
  calculate(stat =  diff in props , order = c( 0 ,  1 ))
## # A tibble: 100 x 2
##    replicate   stat
##        <int>  <dbl>
##  1         1 -0.352
##  2         2 -0.15 
##  3         3 -0.294
##  4         4 -0.254
##  5         5 -0.438
##  6         6 -0.126
##  7         7 -0.188
##  8         8  0.167
##  9         9 -0.143
## 10        10 -0.5  
## # ... with 90 more rows
Two numerical vars - SLR

mtcars %>%
  specify(mpg ~ hp) %>% 
  generate(reps = 100, type =  bootstrap ) %>%
  calculate(stat =  slope )
## # A tibble: 100 x 2
##    replicate    stat
##        <int>   <dbl>
##  1         1 -0.0850
##  2         2 -0.0512
##  3         3 -0.0736
##  4         4 -0.0569
##  5         5 -0.0930
##  6         6 -0.0659
##  7         7 -0.0710
##  8         8 -0.0767
##  9         9 -0.0556
## 10        10 -0.0627
## # ... with 90 more rows
Two numerical vars - correlation

mtcars %>%
  specify(mpg ~ hp) %>% 
  generate(reps = 100, type =  bootstrap ) %>%
  calculate(stat =  correlation )
## # A tibble: 100 x 2
##    replicate   stat
##        <int>  <dbl>
##  1         1 -0.821
##  2         2 -0.812
##  3         3 -0.802
##  4         4 -0.723
##  5         5 -0.885
##  6         6 -0.777
##  7         7 -0.752
##  8         8 -0.758
##  9         9 -0.826
## 10        10 -0.779
## # ... with 90 more rows




Two sample t test example using nycflights13 flights data

https://cran.r-project.org/web/packages/infer/vignettes/two_sample_t.html


Two sample t test example using nycflights13 flights data
Chester Ismay
2018-11-15
Note: The type argument in generate() is automatically filled based on the entries for specify() and hypothesize(). It can be removed throughout the examples that follow. It is left in to reiterate the type of generation process being performed.

Data preparation
library(nycflights13)
library(dplyr)
library(stringr)
library(infer)
set.seed(2017)
fli_small <- flights %>% 
  sample_n(size = 500) %>% 
  mutate(half_year = case_when(
    between(month, 1, 6) ~  h1 ,
    between(month, 7, 12) ~  h2 
  )) %>% 
  mutate(day_hour = case_when(
    between(hour, 1, 12) ~  morning ,
    between(hour, 13, 24) ~  not morning 
  )) %>% 
  select(arr_delay, dep_delay, half_year, 
         day_hour, origin, carrier)
Two numeric - arr_delay, dep_delay
Two categories
half_year ( h1 ,  h2 ),
day_hour ( morning ,  not morning )
Three categories - origin ( EWR ,  JFK ,  LGA )
Sixteen categories - carrier
One numerical variable, one categorical (2 levels)
Calculate observed statistic
The recommended approach is to use specify() %>% calculate():

obs_t <- fli_small %>%
  specify(arr_delay ~ half_year) %>%
  calculate(stat =  t , order = c( h1 ,  h2 ))
## Warning: Removed 15 rows containing missing values.
The observed t statistic is
stat
0.8685
.

Or using t_test in infer

obs_t <- fli_small %>% 
  t_test(formula = arr_delay ~ half_year, alternative =  two_sided ,
         order = c( h1 ,  h2 )) %>% 
  dplyr::pull(statistic)
The observed t statistic is 0.8685.

Or using another shortcut function in infer:

obs_t <- fli_small %>% 
  t_stat(formula = arr_delay ~ half_year, order = c( h1 ,  h2 ))
The observed t statistic is
statistic
0.8685
.

Randomization approach to t-statistic
t_null_perm <- fli_small %>%
  # alt: response = arr_delay, explanatory = half_year
  specify(arr_delay ~ half_year) %>%
  hypothesize(null =  independence ) %>%
  generate(reps = 1000, type =  permute ) %>%
  calculate(stat =  t , order = c( h1 ,  h2 ))
## Warning: Removed 15 rows containing missing values.
visualize(t_null_perm) +
  shade_p_value(obs_stat = obs_t, direction =  two_sided )

Calculate the randomization-based p-value
t_null_perm %>% 
  get_p_value(obs_stat = obs_t, direction =  two_sided )
p_value
0.408
Theoretical distribution
t_null_theor <- fli_small %>%
  # alt: response = arr_delay, explanatory = half_year
  specify(arr_delay ~ half_year) %>%
  hypothesize(null =  independence ) %>%
  # generate() ## Not used for theoretical
  calculate(stat =  t , order = c( h1 ,  h2 ))
## Warning: Removed 15 rows containing missing values.
visualize(t_null_theor, method =  theoretical ) +
  shade_p_value(obs_stat = obs_t, direction =  two_sided )
## Warning: Check to make sure the conditions have been met for the
## theoretical method. {infer} currently does not check these for you.

Overlay appropriate t distribution on top of permuted t-statistics
visualize(t_null_perm, method =  both ) +
  shade_p_value(obs_stat = obs_t, direction =  two_sided )
## Warning: Check to make sure the conditions have been met for the
## theoretical method. {infer} currently does not check these for you.

Compute theoretical p-value
fli_small %>% 
  t_test(formula = arr_delay ~ half_year,
         alternative =  two_sided ,
         order = c( h1 ,  h2 )) %>% 
  dplyr::pull(p_value)
## [1] 0.3855
```