CODEWITHSUNDEEP
rdplyrtidyr
ASF
ASF

Reputation: 279

Using dplyr to filter a dataframe based on a pattern and reorder it

I need to extract all values from all columns in a dataframe based on a pattern (all values with 'EXO' pattern) and reorder it accordingly.

I tried to use a traditional R function and apply, but I get an error related to the different number of rows. I think it would be easier to use a dplyr or tidyr function.

I tried:

df2 = as.data.frame(apply(df,2,function (x) x[grepl("^EXO",x)]))

Error in (function (..., row.names = NULL, check.rows = FALSE, check.names = TRUE, : arguments imply differing number of rows: 2, 1, 7

df <- data.frame(
Group1 = c("EMX1", "EXO_C3L4", "FAF2P1", "FAM224A","NA", 
           "FAM43A", "FAT4", "EXO_FEZF1-AS1"),
Group2 = c("EXO_BRPF3", "NA", "NA", "CCDC187", 
           "CCDC200", "CCDC7", "CCL27", "CD6"),
Group3 = c("SNORD114-18", "SNORD115-10", "SPATA31B1P", "SPIC", 
           "NA", "EXO_TATDN2P3", "EXO_TCIM", "TEPP"),
Group4 = c("EXO_SATB2-AS1", "NA", "EXO_SFTA3", "EXO_SIX3-AS1", "EXO_SMIM2-IT1", 
           "EXO_SNORA46", "EXO_SNORD101", "EXO_SNORD114-18") )

> df
         Group1    Group2       Group3          Group4
1          EMX1 EXO_BRPF3  SNORD114-18   EXO_SATB2-AS1
2      EXO_C3L4        NA  SNORD115-10              NA
3        FAF2P1        NA   SPATA31B1P       EXO_SFTA3
4       FAM224A   CCDC187         SPIC    EXO_SIX3-AS1
5            NA   CCDC200           NA   EXO_SMIM2-IT1
6        FAM43A     CCDC7 EXO_TATDN2P3     EXO_SNORA46
7          FAT4     CCL27     EXO_TCIM    EXO_SNORD101
8 EXO_FEZF1-AS1       CD6         TEPP EXO_SNORD114-18

my expected result, extract only the values with 'EXO' pattern:


> df2
         Group1    Group2       Group3          Group4
1            NA EXO_BRPF3           NA   EXO_SATB2-AS1
2      EXO_C3L4        NA           NA              NA
3            NA        NA           NA       EXO_SFTA3
4            NA        NA           NA    EXO_SIX3-AS1
5            NA        NA           NA   EXO_SMIM2-IT1
6            NA        NA EXO_TATDN2P3     EXO_SNORA46
7            NA        NA     EXO_TCIM    EXO_SNORD101
8 EXO_FEZF1-AS1        NA           NA EXO_SNORD114-18


#And after that, reorder each columns alphabetically:

> df3
         Group1    Group2       Group3          Group4
1      EXO_C3L4 EXO_BRPF3 EXO_TATDN2P3   EXO_SATB2-AS1
2 EXO_FEZF1-AS1        NA     EXO_TCIM       EXO_SFTA3
3            NA        NA           NA    EXO_SIX3-AS1
4            NA        NA           NA   EXO_SMIM2-IT1
5            NA        NA           NA     EXO_SNORA46
6            NA        NA           NA    EXO_SNORD101
7            NA        NA           NA EXO_SNORD114-18
8            NA        NA           NA              NA

Upvotes: 1

Views: 109

Answers (1)

akrun
akrun

Reputation: 886958

Looks like the ordering should be independent. We replace the column values that are not having 'EXO' substring to NA and then do an order

library(dplyr)
library(stringr)
df %>%
    mutate_all(~ replace(., !str_detect(., "EXO"), NA_character_) %>% 
               {.[order(.)]})
#       Group1    Group2       Group3          Group4
#1      EXO_C3L4 EXO_BRPF3 EXO_TATDN2P3   EXO_SATB2-AS1
#2 EXO_FEZF1-AS1      <NA>     EXO_TCIM       EXO_SFTA3
#3          <NA>      <NA>         <NA>    EXO_SIX3-AS1
#4          <NA>      <NA>         <NA>   EXO_SMIM2-IT1
#5          <NA>      <NA>         <NA>     EXO_SNORA46
#6          <NA>      <NA>         <NA>    EXO_SNORD101
#7          <NA>      <NA>         <NA> EXO_SNORD114-18
#8          <NA>      <NA>         <NA>            <NA>

Upvotes: 1

Related Questions