将不规则时间序列数据集与 R 中的规则时间序列数据合并

Question

Hi I am trying to merge two datasets, one with regular time series (dataframe A) and one with irregular time series (dataframe B). The two dataframes look something like:

Dataframe A:

time	country	Action
198001	A	1
198002	A	0
198003	A	0
198004	A	-1
...
201210	Z	1
201211	Z	0
201212	Z	0

Dataframe B:

time	country	party	variable
198002	A	A1	X
201210	Z	Z1	Y
201212	Z	Z2	Z

Ive tried using full_join from dplyr but then it yielded NAs for all timeframes that did not overlap with the observations in dataframe 2.

What I have now looks like:

time	country	Action	Party	Variable
198001	A	1	NA	NA
198002	A	0	A1	X
198003	A	0	A1	NA
198004	A	-1	A1	NA
...
201210	Z	1	Z1	Y
201211	Z	0	Z1	NA
201212	Z	0	Z2	Z

Instead, I would want the observations in the non-overlapping timeframe (NAs) to be replaced with the last observation in dataframe 2.

So the merged dataframe would look like:

time	country	Action	Party	Variable
198001	A	1	NA	NA
198002	A	0	A1	X
198003	A	0	A1	X
198004	A	-1	A1	X
...
201210	Z	1	Z1	Y
201211	Z	0	Z1	Y
201212	Z	0	Z2	Z

Answer 1

You can do a full_join and then fill party and variable columns.

result <- dplyr::full_join(A, B, by = c("time", "country")) %>%
           tidyr::fill(party, variable, .direction = 'down')

result