for about 9285 (38%) of the Site x segment x dates combinations, the perc vegetation does not add up to 100
1) in 132 cases it sums to >100
2) in 7098 cases nothing reported (3 NAs)
3) in 296 cases there is no numeric estimation but an "x" in either category (in all but 1 case only one category is marked)
4) in 1169 cases it sums to less than 100
5) remaining cases include NAs coded as character NA's (the majority), 0 in all categories, and a comment 'Behöver datajustering'
we cannot do much about 2) and 5).
Not sure how to deal with 1) and 4).
For 3) we can replace the field marked with an x by 100.
Note that there are more NAs in cases were the sum is still 100, but there we can assume the NAs code for 0
FabianRoger
commented
3 years ago
for about 9285 (38%) of the Site x segment x dates combinations, the perc vegetation does not add up to 100
1) in 132 cases it sums to >100 2) in 7098 cases nothing reported (3 NAs) 3) in 296 cases there is no numeric estimation but an "x" in either category (in all but 1 case only one category is marked) 4) in 1169 cases it sums to less than 100 5) remaining cases include NAs coded as character NA's (the majority), 0 in all categories, and a comment 'Behöver datajustering'
we cannot do much about 2) and 5). Not sure how to deal with 1) and 4). For 3) we can replace the field marked with an x by 100.
Note that there are more NAs in cases were the sum is still 100, but there we can assume the NAs code for 0