Two-step hybrid refactoring #34
Conversation
maximskorik
commented
Apr 20, 2022
•
maximskorik
commented
- added test data for two-step hybrid
- switched most of data handling functionality to dplyr framework
- refactored two-step hybrid into individual steps
- extracted repeating methods in adjust.time.R and feature.align.R
- reformated semi.sup.R and two.step.hybrid.R
R/two.step.hybrid.R
Outdated
| for (batch_id in batches_idx) | ||
| { | ||
| this.fake <- step_one_features[[batch_id]] | ||
| this.fake.medians <- distinct(this.fake, mz, rt, intensity)$intensity |
There was a problem hiding this comment.
Are these really the medians? If so, how do the distinct intensities form the medians?
There was a problem hiding this comment.
At this step this.fake is a long table where intensity is the column name of median intensities. Since each feature is repeated in multiple rows (a row for a sample) and values in intensity column is the same for all samples of a given feature, distinct selects a median intensity of each unique feature.
In retrospect, calling a column that stores median intensities as intensity is confusing. Fixed that in 976e416.
There was a problem hiding this comment.
cool - how comes that the distinct(...) function actually computes the median intensities? I thought that it actually selects the unique values in the list?
There was a problem hiding this comment.
It doesn't compute the medians. The medians are already computed and stored in the intensity column. distinct(...) only selects the median intensities of each unique feature.
There was a problem hiding this comment.
Refactored this part in 99ae4f5.
R/two.step.hybrid.R
Outdated
| colnames(final.ftrs) <- stringr::str_remove_all(colnames(final.ftrs), ".mzml") | ||
| colnames(final.times) <- stringr::str_remove_all(colnames(final.times), ".mzml") |
There was a problem hiding this comment.
Is this required, to strip the file ending? If so, this should be done more robustly.
There was a problem hiding this comment.
No, it's not required. Having a file extension appended to a sample name in the table just doesn't seem pleasant or valuable.
There was a problem hiding this comment.
Then this should be changed to strip any file ending since there could also be other names than only ".mzml".
| colnames(final.ftrs)[sample_cols_idx] <- tools::file_path_sans_ext(colnames(final.ftrs)[sample_cols_idx]) | ||
| colnames(final.times)[sample_cols_idx] <- tools::file_path_sans_ext(colnames(final.times)[sample_cols_idx]) |
There was a problem hiding this comment.
Nice way to exclude the file endings!
