scaling x and y axis (geom_bar) - r

Question

I am having a trouble plotting what is seemingly simple plot.
x <-
read_excel("Desktop/Book1.xlsx",
col_types = c("numeric", "numeric", "numeric"))
x1 <- gather(hospitals, key = "sector", value = "count", 2:3)
p <- ggplot(data = x1, aes( x = Years, y = count, fill = sector )) +
geom_col(position="stack", stat="identity", width = 5, colour="black") +
geom_text(aes(label=count), vjust=1, color="white", size=2) +
guides(fill=FALSE)+
scale_fill_grey() +
theme_bw(base_size = 12 )
p
data is
1 1946 Public hospitals 35
2 1984 Public hospitals 41
3 2000 Public hospitals 65
4 2001 Public hospitals 67
5 2002 Public hospitals 66
6 2003 Public hospitals 76
7 2004 Public hospitals 77
8 2005 Public hospitals 85
9 2006 Public hospitals 90
10 2007 Public hospitals 94
11 2008 Public hospitals 97
12 2009 Public hospitals 102
13 2010 Public hospitals 102
14 1946 Private hospitals NA
15 1984 Private hospitals 139
16 2000 Private hospitals 325
17 2001 Private hospitals 336
18 2002 Private hospitals 343
19 2003 Private hospitals 364
20 2004 Private hospitals 376
21 2005 Private hospitals 376
22 2006 Private hospitals 353
23 2007 Private hospitals 355
24 2008 Private hospitals 365
25 2009 Private hospitals 370
26 2010 Private hospitals 376
Showing 12 to 26 of 26 entries, 3 total columns
and i am ending with this result!
first, how can I modify x axis to show the bars separated and only for the years i have data for ? [ can the x axis around 1960 be omitted and bars squeezed to save space ?
second, how can the Y axis be fixed ? some bars are higher than their value is!

Answer 1

x1 %>%
ggplot(aes( x = as.character(Years), y = count, fill = sector )) +
geom_col(position="stack", colour="black") +
geom_text(aes(label=count), vjust=1, size=2,
color=ifelse(df$sector != "Public hospitals", "white", "black")) +
guides(fill=FALSE) +
scale_x_discrete(name = "Year") +
scale_fill_grey() +
theme_bw(base_size = 12)
Edit: Upon reconsideration I realized I did not properly stack the positioning of the text. It happens to look ok with this data, but that's just coincidence. To get the right positioning for the text, one approach is manual: we could sum up the cumulative height for each year:
x1 %>%
group_by(Years) %>%
mutate(cuml_count = cumsum(count)) %>%
ungroup() %>% ....
geom_text(aes(label = count, y = cuml_count), vjust = 1, size = 2,
color=ifelse(df$sector != "Public hospitals", "white", "black")) +