如何根据R中另一个dataframe中的列删除列dataframe中的行？

Question

Let's suppose I have two dataframes that look like this:

df1 = structure(list(X1 = c(0.659588465514883, 0.47368422669833, -0.0422047052887636, 
-1.75642936005977, 0.339813114272074, 1.09341750942405, 0.327672990051479, 
-0.893507823167616, -0.661285321563594, -0.569673784617002, -0.983369868281376, 
-2.53659592825309, 0.396220995581641, -1.1994504350227, -0.553343957714012, 
1.30884516680972, -0.120561033997931, 0.971506981390537, 0.815610612704566, 
1.53103368033727, -0.808956975392184, -1.27332589061096, -1.89082047917723, 
0.249755375966669, -0.704051599213331), X2 = c(0.659588465514883, 
0.47368422669833, -0.0422047052887636, -1.75642936005977, 0.339813114272074, 
1.09341750942405, 0.327672990051479, -0.893507823167616, -0.661285321563594, 
-0.569673784617002, -0.983369868281376, -2.53659592825309, 0.396220995581641, 
-1.1994504350227, -0.553343957714012, 1.30884516680972, -0.120561033997931, 
0.971506981390537, 0.815610612704566, 1.53103368033727, -0.808956975392184, 
-1.27332589061096, -1.89082047917723, 0.249755375966669, -0.704051599213331
), Date = structure(c(10957, 
10988, 11017, 11048, 11078, 11109, 11139, 11170, 11201, 11231, 
11262, 11292, 11323, 11354, 11382, 11413, 11443, 11474, 11504, 
11535, 11566, 11596, 11627, 11657, 11688), class = "Date")), class = "data.frame", row.names = c(NA, 
-25L))

            X1           X2
1  -1.633636896 -1.633636896
2   1.793766808  1.793766808
3   0.440697771  0.440697771
4   0.330091148  0.330091148
5  -1.234246285 -1.234246285
6   0.044951993  0.044951993
7  -2.831295687 -2.831295687
8  -0.735371579 -0.735371579
9  -0.412580789 -0.412580789
10  0.001848622  0.001848622
11  1.480684731  1.480684731
12 -1.088999830 -1.088999830
13 -0.465903929 -0.465903929
14 -0.010743010 -0.010743010
15  1.420995930  1.420995930
16 -0.789190729 -0.789190729
17 -0.750476176 -0.750476176
18 -0.314079067 -0.314079067
19 -0.324779959 -0.324779959
20 -1.192471909 -1.192471909
21 -0.170325813 -0.170325813
22  0.890941125  0.890941125
23  0.863875448  0.863875448
24 -0.088048086 -0.088048086
25  0.021239226  0.021239226
    Date
1   2000-01-01
2   2000-02-01
3   2000-03-01
4   2000-04-01
5   2000-05-01
6   2000-06-01
7   2000-07-01
8   2000-08-01
9   2000-09-01
10  2000-10-01
11  2000-11-01
12  2000-12-01
13  2001-01-01
14  2001-02-01
15  2001-03-01
16  2001-04-01
17  2001-05-01
18  2001-06-01
19  2001-07-01
20  2001-08-01
21  2001-09-01
22  2001-10-01
23  2001-11-01
24  2001-12-01
25  2002-01-01

df2 = structure(list(X1 = c(-0.0712460200169048, 1.0131741924359, 0.28590272354409, 
-0.835911047943257, -0.146890264431744), X2 = c(-0.0712460200169048, 
1.0131741924359, 0.28590272354409, -0.835911047943257, -0.146890264431744
), Date = structure(c(10984, 11120, 11441, 11488, 11712), class = "Date")), class = "data.frame", row.names = c(NA, 
-5L))

           X1          X2       Date
1  0.03815189  0.03815189 2000-01-28
2 -0.22665838 -0.22665838 2000-06-12
3  0.36459588  0.36459588 2001-04-29
4  0.32772746  0.32772746 2001-06-15
5 -1.22891784 -1.22891784 2002-01-25

What I would like to do is to reduce the number of rows in df1 (number of rows in df1 = number of rows in df2 ) on the basis of the the number of rows in df2 . In particular, I would like to remove those rows that are in the Date column for df1 is not present in the Date column of df2 . Easier to see the output I would like to get:

# DF1 shall become like this (n stays for the numbers corresponding to each date row):

           X1          X2       Date
1  n                    n 2000-01-01
2  n                    n 2000-06-01
3  n                    n 2001-04-01
4  n                    n 2001-06-01
5  n                    n 2002-01-01

# not really important which day is diplayed in the finale output. What matters is just year and month

I tried to use semin_join but the problem is that different days make the function unable to grasp what I need. Ideally, I would need to ignore days and sample by year and months.

This is what I tried:

library(dplyr)

semin_join(df1, df2, by = "Date")

[1] X1   X2   Date
<0 rows> (or 0-length row.names)

Can anyone help me?

Thanks!

Answer 1

Using the great suggestion from @arg0naut91 here a possible solution in base R . First format the variables Date and then you can use %in% to check which dates are present or not. Next the code using your df1 and df2 :

#Format dates
df1$I1 <- format(df1$Date,'%Y-%m')
df2$I2 <- format(df2$Date,'%Y-%m') 

Now this makes the contrast:

df1[df1$I1 %in% df2$I2,]

Output:

           X1         X2       Date      I1
1   0.6595885  0.6595885 2000-01-01 2000-01
6   1.0934175  1.0934175 2000-06-01 2000-06
16  1.3088452  1.3088452 2001-04-01 2001-04
18  0.9715070  0.9715070 2001-06-01 2001-06
25 -0.7040516 -0.7040516 2002-01-01 2002-01

In the end you could assign that result to a new dataframe and remove I1 .