To compute a dynamic threhsold, change the type option in the compute_threshold function to 'dynamic'. The former will return threshold values that differ for each QC-ART score depending upon the time since the first instrument run, while the latter will return the same threshold value for all QC-ART scores. Otherwise, the single mean model is used. If the SLR model accounts for a statistically significant amount of the variability in the baseline scores (at the 0.05 level) relative to the single mean model, then threshold values from the SLR model are returned. Internally, the compute_threshold function fits two models to the QC-ART scores: a simple log-linear regression model (SLR) where the time since the first instrument run is used as the only covariate, and a single mean model. The column Static_Threshold contains the static threshold value corresponding to each QC-ART score. Qplot(Time_Stamp,Scores,data =sthresh_df,colour =Baseline) + xlab( 'Date') + ylab( 'QC-ART Score') + geom_line( aes(Time_Stamp,Static_Threshold),colour =1) S_threshold <- compute_threshold(scores = vorb_04 $QC_Art, baseline = vorb_04 $Baseline, time_stamps = vorb_04 $Acq_Time_Start, type = 'static')

In the final line of the following, the first six instrument runs from the analysis are removed because they occur well before the rest of the data are collected. For this example we consider the instrument called r chosen_inst, which consists of r length(which(amidan$Instrument=chosen_inst)) instrument runs. We therefore subset the amidan data set to a single instrument before analysis using the dplyr::filter function. The goal of QC-ART is to identify changes in the quality of data produced by MS instruments. The data were collected using r length(unique(amidan$Instument)) different instruments from r round_date(min(amidan$Acq_Time_Start),unit='day') to r round_date(max(amidan$Acq_Time_Start),unit='day'). In short, the amdian data frame consists of r ncol(amidan) variables measured on r nrow(amidan) LC-MS experiments. Now there should be a data frame called amidan loaded into your environment. Amidan $Acq_Time_Start <- mdy_hm(amidan $Acq_Time_Start)

Qcart trial