Package 'blindspiker'

Load data associated with the spiked samples and laboratory results from comma separated variable text files.

Description

This function loads spike data and the laboratory results, then processes it so that the 'sample_ID' ties the spike value and the results together, using the 'left_join' function from the 'dplyr' package.

Usage

bs_prep_and_analysis(spike_data, lab_data)

Arguments

spike_data	name of the loaded dataset (no quotes) or the name of the file, with path, containing the spike values, in quotes. Example forms, my_spike_data, or "C:/my_directory/my_spike_data.csv". This file column headings must be as identified in the Details section.
lab_data	name of the loaded dataset (no quotes) or the name of the file, with path, containing the laboratory results. Example forms, my_lab_data, or "C:/my_directory/my_lab_data.csv". This file column headings must be as identified in the Details section.

Details

False negative results are flagged for laboratory results below the detection level in the analysis of a sample spiked above the detection level. False positives are flagged for laboratory results above the detection when the analyte was not spiked. Error rates are computed with the 'table_false' function.

To use this function, first set up spike value data in a .csv file (in any column order) matching the column headers of the following values:

Required for spike data:

• 'sample_ID' unique identifier, character or numeric

• 'analyte' character data

• 'spike_value' numeric value

• 'spike_units' character data

• 'submission_date' character data that will be converted to date in format YYYY-MM-DD (for example 1999-12-31)

Optional for spike data:

• 'sv_unc' numeric, the uncertainty of the spike value. Default = 0

• 'sv_k' the coverage factor for the spike value uncertainty. Default = 2

• 'provider lab' character name of laboratory providing spiked samples

Required for laboratory results:

• 'sample_ID' must match spike 'sample_ID'

• 'analyte' must match spike 'analyte'

• 'result' numeric value

• 'units' must match 'spike_units'

• 'result_date'

• 'det_lvl' numeric detection level

• 'unc' numeric uncertainty of the laboratory result

#' Optional for lab data:

• 'k' the coverage factor for the result uncertainty. Default = 2

Note that the two data sets (spike values and laboratory results) will be combined by 'sample_ID', and also by 'analyte' if present in both sets. (If the laboratory data includes a non-zero result for an analyte not present in the spike data, that would indicate a false positive.)

Use this function to load the spike data and establish a data frame named 'bs_df'. All plotting and data analysis functions will default to look for this data set.

Value

data frame containing all needed data to be used in subsequent functions.

Examples

example_df <- bs_prep_and_analysis(spike_data = spikevals, lab_data = labvals)

---

Example Laboratory Values

Description

Example Laboratory Values

Usage

labvals

Format

## 'labvals' A data frame with 225 rows and 8 columns:

sample_ID

Unique sample identifier

analyte

what is being analyzed

result_date

date of analysis

k

uncertainty coverage factor

result

value of analysis

unc

measurement uncertainty in result units

units

units of measure

det_lvl

minimum level of detection

---

quantile-quantile plot

Description

QQ plots by isotope

Usage

plot_qq(select_analyte, dat = bs_df)

Arguments

select_analyte	the selected analyte for this run chart
dat	data frame with all data needed as described in 'bs_prep_and_analysis'.Default is 'bs_df'.

Value

quantile-quantile plot of laboratory results of spiked samples

Examples

example_df <- bs_prep_and_analysis(spike_data = spikevals, lab_data = labvals)
plot_qq(select_analyte = 'unknownium', dat = example_df)

---

Plot Run Chart

Description

'plot_run()' produces a Run Chart of the selected analyte. The analyte is selected from the analyte set provided. Results are plotted with error bars (uncertainty with coverage factor of 2) when the result is greater than the detection level. On the ratio version, the uncertainties for the results and spike values are combined as the square root of the sums of the relative uncertainties squared. When original results are plotted, the spike values are shown with a small salmon-colored "+".

Usage

plot_run(
  select_analyte,
  dat = bs_df,
  version = "original",
  log = "n",
  removal_notification = "n"
)

Arguments

select_analyte	the selected analyte for this run chart
dat	data frame with all data needed as described in 'bs_prep_and_analysis'. Default is 'bs_df'.
version	The run chart is either shown with 'original' units, default, or with the result shown as a 'ratio' to the spike value.
log	Set log = "y" to make the y-axis a log scale - original version only. Default is "n".
removal_notification	provide a list of results, by sample_ID that were removed from the current run plot because results were less than or equal to zero. @return run plot of laboratory analyses of spiked samples

Examples

example_df <- bs_prep_and_analysis(spike_data = spikevals, lab_data = labvals)
plot_run(select_analyte = 'unknownium', dat = example_df)

---

Plot turnaround time

Description

Plot time for laboratory analysis by date (result date from lab - spiked sample submitted date)

Usage

plot_tat(select_analyte, dat = bs_df, target_days = 60)

Arguments

select_analyte	the selected analyte for this run chart
dat	data frame with all data needed as described in 'bs_prep_and_analysis'. Default is 'bs_df'.
target_days	The target turnaround time in days. Default = 60.

Value

turn-around-time plot

Examples

example_df <- bs_prep_and_analysis(spike_data = spikevals, lab_data = labvals)
plot_tat(select_analyte = 'unknownium', dat = example_df, target_days = 60)

---

Count combinations of spiked samples

Description

A summary of spiked samples is provided based on combinations of interest.

Usage

spike_combos(analytes, dat = bs_df)

Arguments

analytes	a vector of analytes of interest
dat	data frame with all data needed as described in 'bs_prep_and_analysis'. Default is 'bs_df'.

Value

table of combinations of analytes in blind spikes

Examples

example_df <- bs_prep_and_analysis(spike_data = spikevals, lab_data = labvals)
spike_combos(analytes = c('unknownium', 'Sr-90'), dat = example_df)

---

Example Spike Values

Description

Example Spike Values

Usage

spikevals

Format

## 'spikevals' A data frame with 225 rows and 8 columns:

sample_ID

Unique sample identifier

analyte

what is being analyzed

spike_value

measure of sample spiking

sv_unc

uncertainty spike value

sv_k

coverage factor of uncertainty

spike_unit

unit of measure

units

units of measure

provider_lab

origin of spike sample

submission_date

date provided to lab

---

Make table of false positives and false negatives in laboratory results

Description

Make a table with estimated confidence intervals for false negatives and false positives for an analyte in the data set.

Usage

table_false(select_analyte, dat = bs_df)

Arguments

select_analyte	the selected analyte for this table
dat	data frame with all data needed as described in 'bs_prep_and_analysis'.

Details

False negatives are the number of laboratory results that missed a spiked value. For the false negative rate, the numerator is the number of laboratory results less than detection level for spiked samples. The denominator is the number or spiked samples.

False positives are the number of laboratory results above detection level when the analyte identified by the laboratory was not in the spiked sample. For the false positive rate, the numerator is the number of false positives. The denominator is the number of false positives plus number of true negatives.

The total error rate is the total number of laboratory results with either false negative or false positive results divided by the total number of laboratory results.

Value

table of false positive and false negative results

Examples

example_df <- bs_prep_and_analysis(spike_data = spikevals, lab_data = labvals)
table_false(select_analyte = "Sr-90", dat = example_df)

---

Make table summarizing the blind spikes.

Description

Make a table showing how many blind spike samples contained one or more analytes.

Usage

table_spike(dat = bs_df)

Arguments

dat	data frame with all data needed as described in 'bs_prep_and_analysis'.

Value

table of all analytes in blind spike samples

See Also

For combinations of analytes spiked in a single sample, see 'spike_combos'.

Examples

example_df <- bs_prep_and_analysis(spike_data = spikevals, lab_data = labvals)
table_spike(dat = example_df)

---