This is a test version of Biostars. For the public version, visit https://www.biostars.org.
ttest on a list in R

Hi everyone,

I have a list (data) that I want to apply a t-test on each of the component

[[1]]
         rep_name rep_family   gene distance Hepatocytes_B1 Hepatocytes_B3 Huh.7_B1 Huh.7_B2 T_score
14732 HERVK22-int        LTR THSD7A        0         5.3682           4.74   4.5634   4.0869 0.72895

[[2]]
         rep_name rep_family    gene distance Hepatocytes_B1 Hepatocytes_B3 Huh.7_B1 Huh.7_B2 T_score
2565  HERVL40-int        LTR  ANKIB1     3238         7.2268         7.3056   7.2132   7.5750 -0.1279
2646  HERVL40-int        LTR  ANKIB1     2879         7.2268         7.3056   7.2132   7.5750 -0.1279
2673  HERVL40-int        LTR  ANKIB1     2355         7.2268         7.3056   7.2132   7.5750 -0.1279
2693  HERVL40-int        LTR  ANKIB1     2051         7.2268         7.3056   7.2132   7.5750 -0.1279
16782 HERVL40-int        LTR PRKAR2B        0         6.4382         2.2347   7.6774   6.6859 -2.8452

[[3]]
      rep_name rep_family   gene distance Hepatocytes_B1 Hepatocytes_B3 Huh.7_B1 Huh.7_B2  T_score
990  L1PA15-16       LINE    CFH        0        11.3796        13.2806   6.9655   7.1840  5.25535
1035 L1PA15-16       LINE    CFH        0        11.3796        13.2806   6.9655   7.1840  5.25535
7078 L1PA15-16       LINE KLHL13   152805         5.3824         4.7007   7.0668   6.8522 -1.91795

[[4]]
      rep_name rep_family     gene distance Hepatocytes_B1 Hepatocytes_B3 Huh.7_B1 Huh.7_B2  T_score
11062    LTR13        LTR SLC22A16        0         1.8176         2.4619   3.0775   2.0936 -0.44580
14322    LTR13        LTR   POLR2J      929         7.6368         8.2278   9.5044   9.8754 -1.75760
22284    LTR13        LTR   CACNG3    11594         2.6007         2.4943   1.8240   1.2075  1.03175

[[5]]
      rep_name rep_family   gene distance Hepatocytes_B1 Hepatocytes_B3 Huh.7_B1 Huh.7_B2 T_score
9719    LTR22A        LTR   CD38        0         5.8011         6.9873   5.7066   2.5126 2.28460
15293   LTR22A        LTR THSD7A        0         5.3682         4.7400   4.5634   4.0869 0.72895

the thing that I tried was:

ttest <- for(i in 1:length(data)){
    var1 <- data[[i]][5:6]
    var2 <- data[[i]][7:8]
    t.test(var1,var2)
}

but it gives me an error.However when I try for example to run the test bit by bit(on the second element in the list)

var1 <- data[[2]][5:6]
var2 <- data[[2]][7:8]
t.test(var1,var2)

It works perfectly, does someone has an idea where I m getting it wrong?

Thanks

r
• 6,924 views • link
updated by Ram Ram • written by yasjas • yasjas

Firstly:

What error is thrown?

What is the value of i at the point when the error is thrown?

Are the input values all the same for data[[i]]?

• 1 views • link
updated by Ram Ram • written by russhh russhh

it s just giving me a NULL as answer and for the input I don't know if it s logical but I thought I represents each list like list[[1]],list[[2]],.. etc.

I don't know if I am making sense

• 1 views • link
updated by Ram Ram • written by yasjas • yasjas

It is giving you NULL because for loops don't assign values to variables. You would have to use lapply() for that, or manually build your results list inside the for loop.

• 1 views • link
updated by Ram Ram • written by alolex alolex

Thanks, I will try

• 1 views • link
updated by Ram Ram • written by yasjas • yasjas

1 answer

Try using the lapply() function instead of a for loop. If it still gives an error we would need to know on which list element it gives the error (i.e. which value of i) and what the error is.

lapply(data, function(x){t.test(x[5:6], x[7:8])})
• 1 views • link
updated by Ram Ram • written by alolex alolex

Well I tried this earlier and apparently when there is some values which are constant (compared to the larger of the two means) an error is generated,so it doesn't work either :-(

• 1 views • link
updated by Ram Ram • written by yasjas • yasjas

That error is correct to appear, if the values are identical then the variance is 0, and the t-statistics cannot be calculated. These tests need to be skipped, because the variance estimate is bad.

• 1 views • link
written by Michael Dondrup Michael Dondrup

This is correct. I found this post on how to possibly resolve this issue, but it may not be applicable to your situation, so make sure you understand what you are trying to test before following the suggestion.

• 1 views • link
updated by Ram Ram • written by alolex alolex 
Error in if (stderr < 10 * .Machine$double.eps * max(abs(mx), abs(my))) stop("data are essentially constant") :
  missing value where TRUE/FALSE needed
In addition: Warning messages:
1: In mean.default(x) : argument is not numeric or logical: returning NA
2: In mean.default(y) : argument is not numeric or logical: returning NA

That's the error but I think it s because some of the values are constant in some row

• 1 views • link
updated by Ram Ram • written by yasjas • yasjas

Log in to answer this question.

More posts like this
  • Remove redundant name in a list
    written by yasjas • yasjas
    ``` [[1]] rep_name rep_family gene_name Hepatocytes_B1 Hepatocytes_B3 Huh.7_B1 Huh.7_B2 HERVL40-int LTR ANKIB1 7.2268 7.3056 7.2132 7.5750 HERVL40-int LTR ANKIB1 7.2268 7.3056 7.2132 7.5750 HERVL40-int LTR …
  • t test PROBLEM
    written by yasjas • yasjas
    Hi guys, I have a problem when applying a t-test on my dataframe. I am doing this: ```r class1 <- data[,c(3,5)] # taking the column …
  • boxplot with a for loop
    written by yasjas • yasjas
    Hi all, I tried to make a boxplot for each classes of (`data`) based on repeat names (`rep_name`) with a for loop but it didn't …