Entering edit mode

8.4 years ago

kanwarjag
★
1.2k

I have a data of 9 rows with 25 columns and would l ike to apply logistic regression to subset the data for three grps. Is there a tool which apply logistic regression and then give me subset of my data matrix for 9 rows and columns. I see mostly one gets p value etc. what never a subset of rows of columns which contribute to logistic regression significance.

Thanks

we need a lil more info on what the subset is...first 3 rows? first 3 columns? what are the data? can you expand on that?

It has 9 samples in rows 3 each assigned to one group. There are 3250 (I just wrote 25 in my previous message for simplicity) gene expression observations. The problem is there lot of variations with in group. So I would like to use Logistic regression and extract the subset of three groups (9 samples, rows) with expression in columns.

that still doesn't help...in a logistic regression model y~x you use a continuous variable (x=predictor) to predict a binary outcome (y = 1/0, on/off, yes/no), what are the variables?

you have 9 samples, 3 groups --> 3 groups per sample?

you say that there is lots of variation within single groups, so...you want to extract subsets with lower variation? what are the data? I'm not sure this is the best/correct approach...