从R中两个符号之间的段落中提取文本

Question

I have a paragraph of text and I would like to extract every sample size from it. I usually can get Regex to work, but I am unable to.

Here is an example:

x = "OBJECTIVES:

In diabetic polyneuropathy (DPN) patients, the effect of folic acid and homocysteine has been related to components of nerve conduction velocity (NCV). The objective of this study was to determine the effect of folic acid supplementation on NCV in DPN patients.
METHODS:

Patients were randomized to receive either 1 mg of folic acid (n = 40) or placebo (n = 40) for 16 weeks. Blood samples were collected to assess serum folic acid and homocysteine concentrations, and NCV was performed for assessment of diabetic neuropathy.
RESULTS:

At 16 weeks, in the supplemented group, serum levels of folic acid (p < 0.001) increased, homocysteine concentrations decreased (p < 0.001), with no change in serum vitamin B12 levels. There was a significant increase in sensory sural amplitude (p < 0.001), and components of motor nerves, including amplitude (p = 0.001) and velocity (p < 0.001), but decreased onset latency of peroneal (p = 0.019) and tibial (p = 0.011) motor nerves.
CONCLUSION:

Our data suggest that supplementation with 1 mg of folic acid for 16 weeks may be useful for enhancing NCV in DPN patients."

I would like to extract out the two sample sizes. In this case n = 40 and n = 40.

I have tried

gsub('.*[n=]|).*','',x)

I get back ts.

Answer 1

Here's one way to extract those values

regmatches(x, gregexpr('n\\s*=\\s*\\d+',x))

here we look for n= (with possible spaces around the equals sign) and then extract those with the regmatches .

Answer 2

An ugly solution without regex could be:

#first "n = "
substr(strsplit(x, "n = ",fixed=T)[[1]][2],1,2)
#second "n = "
substr(strsplit(x, "n = ",fixed=T)[[1]][3],1,2)

Answer 3

You could use stringr to extract "n = " followed by at least one digit. This assumes there will be no-space or one space either side of the equal sign:

library(stringr)
str_extract_all(x, "n\\s?\\=\\s?\\d+")
[[1]]
[1] "n = 40" "n = 40"

EDIT: The following should work inside mutate with your other condition. I switched from stringr to stringi to get NA for a string with no matches. Also, you could use paste instead of stri_flatten but I would stick with stri_flatten because it retains NA as a missing value and not a character "NA" like paste does.

sapply(stri_extract_all(x, regex = "n\\s?\\=\\s?\\d+"), stri_flatten, collapse = ", ")

For regex I started with this cheat sheet for R (and still reference it). The above regex works like so:

n - the letter n

\\\\s? - at most 1(the ? ) space( \\\\s ) (you may prefer MrFlick's use of * over the ? --your call)

\\\\= - equal sign

\\\\s? - at most 1(the ? ) space( \\\\s )

\\\\d+ - one or more( + ) digits( \\\\d )

Data :

x = c("OBJECTIVES:

In diabetic polyneuropathy (DPN) patients, the effect of folic acid and homocysteine has been related to components of nerve conduction velocity (NCV). The objective of this study was to determine the effect of folic acid supplementation on NCV in DPN patients.
METHODS:

Patients were randomized to receive either 1 mg of folic acid (n = 40) or placebo (n = 40) for 16 weeks. Blood samples were collected to assess serum folic acid and homocysteine concentrations, and NCV was performed for assessment of diabetic neuropathy.
RESULTS:

At 16 weeks, in the supplemented group, serum levels of folic acid (p < 0.001) increased, homocysteine concentrations decreased (p < 0.001), with no change in serum vitamin B12 levels. There was a significant increase in sensory sural amplitude (p < 0.001), and components of motor nerves, including amplitude (p = 0.001) and velocity (p < 0.001), but decreased onset latency of peroneal (p = 0.019) and tibial (p = 0.011) motor nerves.
CONCLUSION:

Our data suggest that supplementation with 1 mg of folic acid for 16 weeks may be useful for enhancing NCV in DPN patients.", "no numbers here", "n = 100")

Answer 4

A way to retrive the number between parenthesis is like this:

library(stringr)

lapply(str_split(x,pattern="\\("),function(x) gsub('(.*)\\).*','\\1',x))
[[1]]
 [1] "OBJECTIVES:\n\nIn diabetic polyneuropathy "
 [2] "DPN"                                       
 [3] "NCV"                                       
 [4] "n = 40"                                    
 [5] "n = 40"                                    
 [6] "p < 0.001"                                 
 [7] "p < 0.001"                                 
 [8] "p < 0.001"                                 
 [9] "p = 0.001"                                 
[10] "p < 0.001"                                 
[11] "p = 0.019"                                 
[12] "p = 0.011"    

You split the text using \\( as a pattern and apply the gsub in each piece of it. Afterwards you can use grep to identify which elements start with "n =" to retrieve the one you need.

I hope it helps