0

Here is my homework:

In a first step, describe the dataset as well as the variables contained within it. Use descriptive measures and figures to do this. Then, compare Y T =1 to Y T =0 in a second step. Use an appropriate figure to visualize the relationship between these variables. Finally, use the data to argue whether or not it is likely that treatment is randomly assigned in a third step.

Here is my code:

library(tidyverse)
library(readr)
X17856731 <- read_csv("17856731.csv")
View(X17856731)

df_X21751433<-read_csv("17856731.csv")

mean(X17856731$X)
mean(X17856731$Y)

Mode = function(x, na.rm = FALSE)  { 
  if(na.rm) { 
  x = na.omit(x) 
}
ux = unique(x)
return(ux[which.max(tabulate(match(x)))])
}

Mode(X17856731$X, na.rm = TRUE)
Mode(X17856731$Y, na.rm = TRUE)

median(X17856731$X, na.rm = TRUE)
median(X17856731$Y, na.rm = TRUE)

var(X17856731$X, na.rm = TRUE)
var(X17856731$Y, na.rm = TRUE)

sd(X17856731$X, na.rm = TRUE)
sd(X17856731$Y, na.rm = TRUE)

install.packages("moments")
library(moments)
skewness(X17856731$X, na.rm = TRUE)
skewness(X17856731$Y, na.rm = TRUE)

IQR(X17856731$X, na.rm = TRUE)
IQR(X17856731$Y, na.rm = TRUE)

df_nottreated_y = filter(X17856731, T == 0)
df_treated_y = filter(X17856731, T == 1)

mean(df_treated_y$Y, na.rm = TRUE) - mean(df_nottreated_y$Y, na.rm = TRUE)

X17856731%>%
  group_by(T) %>%
  summarize(df_nottreated_y = mean(Y),
            df_treated_y = mean(Y))

select(X17856731, Y)
head(
  select(X17856731, Y, T = 0)
)

select(X17856731, Y)
head(
  select(X17856731, Y , T = 1)
)

df_selects = X17856731 :: selects %>%
  select(Y) %>%
  mutate(
    Y= recode(treatment,
             "treated" = 1,
             not_treated = 0)
  ) %>% na.omit
df_selects

data(treated_y, package = 'dplyr')
treated_y

data(nottreated_y, package = 'dplyr')
nottreated_y
halfer
  • 19,824
  • 17
  • 99
  • 186
tatimaibo
  • 1
  • From the online doc for `match`: "match returns a vector of the positions of (first) matches of its first argument in its second." You've only provided a single argument: `return(ux[which.max(tabulate(match(x)))])`. Though I get a different error to you: `Error in match(x) : argument "table" is missing, with no default`. `base::mode` is probably an easier way of identifying the mode. – Limey Oct 23 '22 at 10:09
  • It looks like you have posted your homework problem into the question, but this does not really bear any relation to the title of the question. Could you edit the question to clarify exactly what help you need? – halfer Oct 23 '22 at 11:23

0 Answers0