tkimhofer / metabom8 Goto Github PK

R library for NMR/MS analyses. Workflow includes NMR data import & preprocessing, statistical modelling (PCA, O-PLS), met ID-ing, visualisation (ggplot2, plotly).. Critical functions implemented in C++ using linear algebra packages (Armadillo, Eigen) to improve run-time.

Home Page: https://tkimhofer.github.io/metabom8/

License: Other

C++ 12.52% R 87.48%

rcpp armadillo eigen

metabom8's Introduction

Hi there 👋

metabom8's People

Contributors

Stargazers

Watchers

Forkers

josoriom reika2 rgilredondo wenjun1992 kanok121212

metabom8's Issues

Glucose calibration error of plasma

Xc=calibrate(Xnew,ppm, type='glucose')
Error in FUN(X[[i]], ...) : object 'crit' not found

read1d: No files found that match the specified parameter specification levels.

trying to read in raspberry NMR data with both read1d and read1d_raw and saying files don't match specified parameters even though there are valid data in there

noise estimation works for NMR matrix but not for individual spectrum

specify error

specify error when using plot load with non NMR data

plotStocsy error

provide a stocsy object error

fix colour argument for sepctral overlay plot with superimposed reconstructed loadings

Binning in pqn

pqn with binning = NULL throws error can't find X1

Also, implement noise cutoff to exclude noise reagion

Importing non-IVDr NMR spectra

path='../../../20190330/Onedrive - Murdoch University/Rfolder_data/Cambridge_DIRE_280820/'

read1d(path)
Error in .filterExp_files(pars, exp_type, f_list) :
No files found that match the specified parameter specification levels.

NMR: PCA or OPLS model loadings: add cor(t,X) in addition to backscaling with SD

n_max argument does not consider experiment type

testing i can log an issue

test1

opls warning

smod=opls(X,Y)
Performing discriminant analysis.
Setting CV k-fold parameter to 7. Number of observations in group 2 is too low (n=7) for k=7.
An O-PLS-DA model with 1 predictive and 1 orthogonal components was fitted.
Warning messages:
1: In roc.default(response = Y, predictor = preds_test, quiet = TRUE) :
Deprecated use a matrix as predictor. Unexpected results may be produced, please pass a numeric vector.
2: In roc.default(response = Y, predictor = preds_test, quiet = TRUE) :
Deprecated use a matrix as predictor. Unexpected results may be produced, please pass a numeric vector.