Generate a bar graph using ggplot2 and dplyr in R

You have several issues in your code.

1) When you use summarise, your column education.num will be replaced by mean in your dataset as you can see below:

library(dplyr)

mydata %>%
  group_by(workclass) %>%
  summarise(mean = mean(education.num, na.rm = TRUE))

# A tibble: 9 x 2
  workclass         mean
  <chr>            <dbl>
1 ?                 9.26
2 Federal-gov      11.0 
3 Local-gov        11.0 
4 Never-worked      7.43
5 Private           9.88
6 Self-emp-inc     11.1 
7 Self-emp-not-inc 10.2 
8 State-gov        11.4 
9 Without-pay       9.07

2) then, in your ggplot, you are calling an another dataframe new_data and try to reuse education.num instead of mean. You can correct it by doing:

library(dplyr)
library(ggplot2)

mydata %>%
  group_by(workclass) %>%
  summarise(mean = mean(education.num, na.rm = TRUE)) %>% 
  ggplot(aes(workclass, mean, fill = workclass)) +
  geom_bar(stat = "identity") +
  labs(title = "Average Education Num vs workclass",
       x = "Workclass",
       y = "Average Education Num") +
  theme(axis.text.x = element_text(size = 10, angle = 90, hjust = 1),
        axis.text = element_text(size = 10),
        axis.title = element_text(size = 10), 
        legend.title = element_text(size = 10)) 

3) Finally, you are trying to replace filling values by cbPalette, however, you are providing only 8 values whereas you have 9 different classes, so you need to either add a new color and remove ? like that:

library(dplyr)
library(ggplot2)

mydata %>%
  group_by(workclass) %>%
  summarise(mean = mean(education.num, na.rm = TRUE)) %>% 
  filter(workclass != "?") %>%
  ggplot(aes(workclass, mean, fill = workclass)) +
  geom_bar(stat = "identity") +
  labs(title = "Average Education Num vs workclass",
       x = "Workclass",
       y = "Average Education Num") +
  theme(axis.text.x = element_text(size = 10, angle = 90, hjust = 1),
        axis.text = element_text(size = 10),
        axis.title = element_text(size = 10), 
        legend.title = element_text(size = 10)) +
  scale_fill_manual(values = cbPalette) 

Does it answer your question ?