CODEWITHSUNDEEP
rdataframesubset
brucezepplin
brucezepplin

Reputation: 9792

subset dataframe by column groupings

I have the following dataframe:

            V1 subst_string
1     NM_000171        L374R
2     NM_000171        W421P
3     NM_000171        T358A
4     NM_000171        T358A
5     NM_000171        T358A
6     NM_000171        T358A
7     NM_000171        S268R
8     NM_000171         N35P
9     NM_000171        T435F
10    NM_000171        T435F
11    NM_000171        T435F
12    NM_000171        T435F
13    NM_000171        L368E
14    NM_000171        L368G
15    NM_000171        L374R
16    NM_000171        W421L
17    NM_000171        W421P
18    NM_000171        W421L
19    NM_000171        R371R
20    NM_000171        R371R
21    NM_000171        S268R
22    NM_000171         N35R
23    NM_000171         N35P
24    NM_000171         N35R
25    NM_000171        F271L
26    NM_000171        F271L
27    NM_000171        F271L
28    NM_000171        L368E
29    NM_000171        L374R
30    NM_000171        L374R
31    NM_000171        M157N
32    NM_000171        M157N
33    NM_000171        M157N
34    NM_000171        R371R
35    NM_000171        S268R
36    NM_000171        S268R
37    NM_000171        Y201P
38    NM_000171        Y201P
39    NM_000171        Y201P
40    NM_000171        Y201P
41    NM_000171        F271L
42    NM_000171        L368G
43    NM_000171        Y397S
44    NM_000171        Y397G
45    NM_000171        Y397S
46    NM_000171        Y397G
47    NM_000171        M157N
48    NM_000171        R371R
49 NM_001146040        F271L
50 NM_001146040        L368E
51 NM_001146040        L374R
52 NM_001146040        E429P
53 NM_001146040        T358A
54 NM_001146040        T358A
55 NM_001146040        M157N
56 NM_001146040        R371R
57 NM_001146040        S268R
58 NM_001146040         N35P
59 NM_001146040        I443F
60 NM_001146040        I443F
61 NM_001146040        Y201P
62 NM_001146040        Y201P
63 NM_001146040        F271L
64 NM_001146040        L368G
65 NM_001146040        L374R
66 NM_001146040        E429L
67 NM_001146040        L405S
68 NM_001146040        L405G
69 NM_001146040        M157N
70 NM_001146040        R371R
71 NM_001146040        S268R
72 NM_001146040         N35R
73 NM_001292000       NANANA
74 XM_005268412       NANANA

which is fairly simple - the first column consists of 4 distinct values, and I want to create 4 subsets of the data, where the dataframe is subsetted by these distinct values. i.e.

            V1 subst_string
1     NM_000171        L374R
2     NM_000171        W421P
3     NM_000171        T358A
4     NM_000171        T358A
5     NM_000171        T358A
6     NM_000171        T358A
7     NM_000171        S268R
8     NM_000171         N35P
9     NM_000171        T435F
10    NM_000171        T435F
11    NM_000171        T435F
12    NM_000171        T435F
13    NM_000171        L368E
14    NM_000171        L368G
15    NM_000171        L374R
16    NM_000171        W421L
17    NM_000171        W421P
18    NM_000171        W421L
19    NM_000171        R371R
20    NM_000171        R371R
21    NM_000171        S268R
22    NM_000171         N35R
23    NM_000171         N35P
24    NM_000171         N35R
25    NM_000171        F271L
26    NM_000171        F271L
27    NM_000171        F271L
28    NM_000171        L368E
29    NM_000171        L374R
30    NM_000171        L374R
31    NM_000171        M157N
32    NM_000171        M157N
33    NM_000171        M157N
34    NM_000171        R371R
35    NM_000171        S268R
36    NM_000171        S268R
37    NM_000171        Y201P
38    NM_000171        Y201P
39    NM_000171        Y201P
40    NM_000171        Y201P
41    NM_000171        F271L
42    NM_000171        L368G
43    NM_000171        Y397S
44    NM_000171        Y397G
45    NM_000171        Y397S
46    NM_000171        Y397G
47    NM_000171        M157N
48    NM_000171        R371R

would be one such subset. I suppose I could do this manuall, however I envisage far more distinct values for V1 in later datasets. Does anyone know how I can automatically generate subsets of this dataframe, based on distinct values for V1?

Thank you

Upvotes: 0

Views: 102

Answers (2)

rnso
rnso

Reputation: 24623

If you know what to do after you split, there are many methods; and you do not need to use 'split' for that. For example to find length of subgroups: 

with(ddf, tapply(subst_string, V1, length))
   NM_000171 NM_001146040 NM_001292000 XM_005268412 
          48           24            1            1 

aggregate(subst_string~V1, data=ddf, length)
            V1 subst_string
1    NM_000171           48
2 NM_001146040           24
3 NM_001292000            1
4 XM_005268412            1

library(data.table)
ddt = data.table(ddf)
ddt[,list(len =.N),by=V1]
             V1 len
1:    NM_000171  48
2: NM_001146040  24
3: NM_001292000   1
4: XM_005268412   1

Upvotes: 3

akrun
akrun

Reputation: 887851

Just use split to create a list

 split(df, df$V1)

If you need 4 datasets in the global environment

 list2env(split(df, df$V1), envir=.GlobalEnv)

 head(NM_000171,2)
 #        V1 subst_string
 #1 NM_000171        L374R
 #2 NM_000171        W421P

Upvotes: 4

Related Questions