Subdividing and counting how many values in particular columns under certain conditions in r - r

Question

I am new to R and data analysis. I have a database similar to this one below, just a lot bigger and I was trying to find a general way to count for each country how many actions there are and how many subquestion with value 1, value 2 and so on there are. For each action there are multiple questions, subquestions and subsubquestions but I would love to find a way to count
1:how many actions there are per country, excluding subquestions
2: a way to find out how many subquestions 1 or 2 with value 1 there are for each country, actionn and questionn.
id country questionn subquestion value actionn
06 NIE 1 1 1 1
05 NIG 1 1 1 1
07 TAN 1 1 1 1
08 BEN 1 1 1 1
03 TOG 1 1 2 1
45 MOZ 1 1 2 1
40 ZIM 1 1 1 1
56 COD 1 1 1 1
87 BFA 1 1 1 1
09 IVC 1 1 2 1
08 SOA 1 1 2 1
02 MAL 1 1 2 1
78 MAI 1 1 2 1
35 GUB 1 1 2 1
87 RWA 1 1 2 1
41 ETH 1 1 1 1
06 NIE 1 2 2 1
05 NIG 1 2 1 1
87 BFA 1 2 1 2
I have tried to create subsets of the data frame and count everything for each country once at a time but it is going to take forever and I was wondering if there was a general way to do it.
For the first question I have done this
df1<-df %>% group_by (country) %>% summarise (countries=county)
unique(df1)
count(df1)
For the second question I was thinking of individually select and count each rows which has quesionn=1, subquestion=1, value=1 and actionn=1, then to select and count how many per country with qustionn=1, subquestionn=2, value=1, actionn=1 etc. Value refers to whether the answer to the question is 1=yes or 2=no.
I would be grateful for any help, thank you soo much :)

Answer 1

For the first question you can try to do something like this:
df %>%
filter(subquestion != 2) %>%
group_by(country) %>%
summarise(num_actions = n())
This will return the number of actions per country, removing rows that do not have 2 for the subquestion column. Note that the n() in the summarize function will count the number observations in the groups (in this case countries).
I am not sure I fully understand the second question, but my suggestion would be to make a new label for the particular observation you want to know (how many subquestions 1 or 2 with value 1 there are for each country, actionn and questionn):
df %>%
mutate(country_question_code = paste(country, action, questionn, sep = "_")) %>%
group_by(country_question_code) %>%
summarize(num_subquestion = n())

Answer 2

For question 1 possible solution (assuming country name are not unique and actionn can be 0, 1,2 or more..
For just total count:
df%>%group_by(country)%>%
summarise(
"Count_actions" = sum(actionn)
) #ignores all other columns.
If you want to count how many times a country appears use n() in place of sum(actionn, na.rm=TRUE).# this may not be desired but sometime simple solution is the best
(Just count the frequency of country)
Or df%>%group_by(country, actionn)%>%summarise("count_actions"= n()) will give country wise count for each type ( say 1,2 or more actions).
Data table version dt[, .(.N), by=.(country, actionn )]
For question 2: use grouping for "for each on your question" after putting filter on data as required. Here, filter subquestions 1 or 2 with (and) value 1 for each "country, question and actionn":
df%>%filter(subquestions <=2 & value==1)%>%group_by( country, question, actionn)%>%summarise("counts_desired"= n(), "sums_desired"= sum(actionn, na.rm=TRUE))
Hope this works. I am also learning and applying it on similar data.
Have not tested it and made certain assumptions about your data (numerical and clean). (Also for.mobile while traveling! Cheers!!)