Hi all,
May I know if this warning message affect downstream analysis? Couldn't find this error after googling. Thank you so much!
design <- model.matrix(~ vector_age + sex_vector + group_vector, compare_genes)
fit <- lmFit(numeric_matrix, design)
fit2 <- contrasts.fit(fit, contrast_matrix)
Warning message:
In contrasts.fit(fit, contrast_matrix) :
row names of contrasts don't match col names of coefficients
The message seems self-explanatory. It is telling you that row.names(contrast_matrix) is not the same as colnames(fit).
Like all warning messages in R, it doesn't say that the analysis is necessarily wrong (that would be an error rather than an warning) but it prompts you that something looks strange and that you should double-check that what you are doing is correct. Does the contrast matrix that you have formed correctly correspond to this design matrix or are you inputing the wrong contrast matrix?
Use of this site constitutes acceptance of our User Agreement and Privacy Policy.
version 2.3.6
Hi Gordon,
My contract matrix has rows name: Intercept, vector_age, sex_vector2, group_vectorMCI. Column name design matrix: (Intercept), vector_age, sex_vector2, group_vectorMCI. I think the warning because Intercept in contract matrix doesn't have () as in design matrix. So we can ignore this warning, right? Thank you!
That's right, you can ignore the warning.