diff --git a/DESCRIPTION b/DESCRIPTION
index b95d2456..af59cfc0 100644
--- a/DESCRIPTION
+++ b/DESCRIPTION
@@ -1,6 +1,6 @@
Package: AMR
Version: 0.5.0.9017
-Date: 2019-02-08
+Date: 2019-02-09
Title: Antimicrobial Resistance Analysis
Authors@R: c(
person(
diff --git a/NEWS.md b/NEWS.md
index 16b6a055..096fbcc9 100755
--- a/NEWS.md
+++ b/NEWS.md
@@ -87,6 +87,7 @@
* New colours for `scale_rsi_colours()`
* Summaries of class `mo` will now return the top 3 and the unique count, e.g. using `summary(mo)`
* Small text updates to summaries of class `rsi` and `mic`
+* Function `as.rsi()` now gives a warning when inputting MIC values
* Frequency tables (`freq()` function):
* Support for tidyverse quasiquotation! Now you can create frequency tables of function outcomes:
```r
diff --git a/R/resistance_predict.R b/R/resistance_predict.R
index caf0d217..2380b533 100755
--- a/R/resistance_predict.R
+++ b/R/resistance_predict.R
@@ -30,11 +30,17 @@
#' @param year_max highest year to use in the prediction model, defaults to 10 years after today
#' @param year_every unit of sequence between lowest year found in the data and \code{year_max}
#' @param minimum minimal amount of available isolates per year to include. Years containing less observations will be estimated by the model.
-#' @param model the statistical model of choice. Valid values are \code{"binomial"} (or \code{"binom"} or \code{"logit"}) or \code{"loglin"} (or \code{"poisson"}) or \code{"linear"} (or \code{"lin"}).
+#' @param model the statistical model of choice. Defaults to a generalised linear regression model with binomial distribution, assuming that a period of zero resistance was followed by a period of increasing resistance leading slowly to more and more resistance. See Details for valid options.
#' @param I_as_R a logical to indicate whether values \code{I} should be treated as \code{R}
#' @param preserve_measurements a logical to indicate whether predictions of years that are actually available in the data should be overwritten by the original data. The standard errors of those years will be \code{NA}.
#' @param info a logical to indicate whether textual analysis should be printed with the name and \code{\link{summary}} of the statistical model.
#' @param main title of the plot
+#' @details Valid options for the statistical model are:
+#' \itemize{
+#' \item{\code{"binomial"} or \code{"binom"} or \code{"logit"}: a generalised linear regression model with binomial distribution}
+#' \item{\code{"loglin"} or \code{"poisson"}: a generalised log-linear regression model with poisson distribution}
+#' \item{\code{"lin"} or \code{"linear"}: a linear regression model}
+#' }
#' @return \code{data.frame} with extra class \code{"resistance_predict"} with columns:
#' \itemize{
#' \item{\code{year}}
@@ -306,7 +312,9 @@ plot.resistance_predict <- function(x, main = paste("Resistance prediction of",
ylab = paste0("Percentage (", ylab, ")"),
xlab = "Year",
main = main,
- sub = paste0("(model: ", attributes(x)$model_title, ")"))
+ sub = paste0("(n = ", sum(x$observations, na.rm = TRUE),
+ ", model: ", attributes(x)$model_title, ")"),
+ cex.sub = 0.75)
axis(side = 2, at = seq(0, 1, 0.1), labels = paste0(0:10 * 10, "%"))
@@ -332,12 +340,13 @@ ggplot_rsi_predict <- function(x, main = paste("Resistance prediction of", attri
}
suppressWarnings(
ggplot2::ggplot(x, ggplot2::aes(x = year, y = value)) +
- ggplot2::geom_col() +
- ggplot2::geom_errorbar(ggplot2::aes(ymin = se_min, ymax = se_max)) +
- scale_y_percent() +
+ ggplot2::geom_point(size = 2) +
+ ggplot2::geom_errorbar(ggplot2::aes(ymin = se_min, ymax = se_max), na.rm = TRUE, width = 0.5) +
+ scale_y_percent(limits = c(0, 1)) +
ggplot2::labs(title = main,
y = paste0("Percentage (", ylab, ")"),
x = "Year",
- caption = paste0("(model: ", attributes(x)$model_title, ")"))
+ caption = paste0("(n = ", sum(x$observations, na.rm = TRUE),
+ ", model: ", attributes(x)$model_title, ")"))
)
}
diff --git a/R/rsi.R b/R/rsi.R
index 7f8f4fc7..12b17b67 100755
--- a/R/rsi.R
+++ b/R/rsi.R
@@ -64,6 +64,9 @@ as.rsi <- function(x) {
} else if (identical(levels(x), c("S", "I", "R"))) {
structure(x, class = c('rsi', 'ordered', 'factor'))
} else {
+ if (mic_like(x) > 0.5) {
+ warning("`as.rsi` is intended to clean antimicrobial interpretations - not to interpret MIC values.", call. = FALSE)
+ }
x <- x %>% unlist()
x.bak <- x
@@ -103,6 +106,14 @@ as.rsi <- function(x) {
}
}
+mic_like <- function(x) {
+ mic <- x %>%
+ gsub("[^0-9.,]+", "", .) %>%
+ unique()
+ mic_valid <- suppressWarnings(as.mic(mic))
+ sum(!is.na(mic_valid)) / length(mic)
+}
+
#' @rdname as.rsi
#' @export
is.rsi <- function(x) {
diff --git a/README.md b/README.md
index 73a6eb87..9415d705 100755
--- a/README.md
+++ b/README.md
@@ -54,7 +54,7 @@ This is the latest **development version**. Although it may contain bugfixes and
Development Test | Result | Reference
--- | :---: | ---
-All functions checked on Linux | [](https://gitlab.com/msberends/AMR/commits/master) | GitLab CI [[ref 1]](https://gitlab.com/msberends/AMR/pipelines)
+All functions checked on Linux | [](https://gitlab.com/msberends/AMR/commits/master) | GitLab CI [[ref 1]](https://gitlab.com/msberends/AMR)
All functions checked on Windows | [](https://ci.appveyor.com/project/msberends/amr-svxon) | Appveyor Systems Inc. [[ref 2]](https://ci.appveyor.com/project/msberends/amr-svxon)
Percentage of syntax lines checked | [](https://codecov.io/gl/msberends/AMR) | Codecov LLC [[ref 3]](https://codecov.io/gl/msberends/AMR)
diff --git a/docs/articles/AMR.html b/docs/articles/AMR.html
index bbce516b..66955ce3 100644
--- a/docs/articles/AMR.html
+++ b/docs/articles/AMR.html
@@ -40,7 +40,7 @@
AMR (for R)
- 0.5.0.9016
+ 0.5.0.9017
@@ -185,7 +185,7 @@
How to conduct AMR analysis
Matthijs S. Berends
- 08 February 2019
+ 09 February 2019
AMR.Rmd
@@ -194,7 +194,7 @@
-Note: values on this page will change with every website update since they are based on randomly created values and the page was written in RMarkdown. However, the methodology remains unchanged. This page was generated on 08 February 2019.
+Note: values on this page will change with every website update since they are based on randomly created values and the page was written in RMarkdown. However, the methodology remains unchanged. This page was generated on 09 February 2019.
Introduction
@@ -210,21 +210,21 @@
-| 2019-02-08 |
+2019-02-09 |
abcd |
Escherichia coli |
S |
S |
-| 2019-02-08 |
+2019-02-09 |
abcd |
Escherichia coli |
S |
R |
-| 2019-02-08 |
+2019-02-09 |
efgh |
Escherichia coli |
R |
@@ -313,70 +313,70 @@
-| 2010-10-06 |
-F7 |
+2012-10-05 |
+F3 |
Hospital D |
-Klebsiella pneumoniae |
-S |
+Escherichia coli |
+R |
S |
S |
S |
M |
-| 2015-07-29 |
-T1 |
+2012-11-18 |
+X7 |
Hospital A |
Escherichia coli |
-R |
+S |
S |
S |
S |
F |
-| 2017-10-20 |
-P2 |
+2012-04-27 |
+K2 |
Hospital B |
-Staphylococcus aureus |
-S |
+Escherichia coli |
+R |
S |
R |
S |
+M |
+
+
+| 2014-09-03 |
+M6 |
+Hospital A |
+Streptococcus pneumoniae |
+S |
+S |
+S |
+S |
+M |
+
+
+| 2015-09-13 |
+R4 |
+Hospital A |
+Escherichia coli |
+S |
+S |
+S |
+S |
F |
-| 2010-02-07 |
-Z6 |
-Hospital A |
-Klebsiella pneumoniae |
+2011-09-19 |
+D6 |
+Hospital B |
+Streptococcus pneumoniae |
+S |
I |
-R |
S |
S |
-F |
-
-
-| 2012-06-26 |
-V4 |
-Hospital A |
-Staphylococcus aureus |
-S |
-S |
-S |
-S |
-F |
-
-
-| 2016-02-08 |
-S2 |
-Hospital B |
-Escherichia coli |
-S |
-S |
-S |
-S |
-F |
+M |
@@ -396,8 +396,8 @@
#
# Item Count Percent Cum. Count Cum. Percent
# --- ----- ------ -------- ----------- -------------
-# 1 M 2,551 51.0% 2,551 51.0%
-# 2 F 2,449 49.0% 5,000 100.0%
+# 1 M 2,565 51.3% 2,565 51.3%
+# 2 F 2,435 48.7% 5,000 100.0%
So, we can draw at least two conclusions immediately. From a data scientist perspective, the data looks clean: only values M and F. From a researcher perspective: there are slightly more men. Nothing we didn’t already know.
The data is already quite clean, but we still need to transform some variables. The bacteria column now consists of text, and we want to add more variables based on microbial IDs later on. So, we will transform this column to valid IDs. The mutate() function of the dplyr package makes this really easy:
data <- data %>%
@@ -428,10 +428,10 @@
# Kingella kingae (no changes)
#
# EUCAST Expert Rules, Intrinsic Resistance and Exceptional Phenotypes (v3.1, 2016)
-# Table 1: Intrinsic resistance in Enterobacteriaceae (345 changes)
+# Table 1: Intrinsic resistance in Enterobacteriaceae (306 changes)
# Table 2: Intrinsic resistance in non-fermentative Gram-negative bacteria (no changes)
# Table 3: Intrinsic resistance in other Gram-negative bacteria (no changes)
-# Table 4: Intrinsic resistance in Gram-positive bacteria (673 changes)
+# Table 4: Intrinsic resistance in Gram-positive bacteria (681 changes)
# Table 8: Interpretive rules for B-lactam agents and Gram-positive cocci (no changes)
# Table 9: Interpretive rules for B-lactam agents and Gram-negative rods (no changes)
# Table 10: Interpretive rules for B-lactam agents and other Gram-negative bacteria (no changes)
@@ -447,9 +447,9 @@
# Non-EUCAST: piperacillin/tazobactam = S where piperacillin = S (no changes)
# Non-EUCAST: trimethoprim/sulfa = S where trimethoprim = S (no changes)
#
-# => EUCAST rules affected 1,814 out of 5,000 rows
+# => EUCAST rules affected 1,796 out of 5,000 rows
# -> added 0 test results
-# -> changed 1,018 test results (0 to S; 0 to I; 1,018 to R)
+
# -> changed 987 test results (0 to S; 0 to I; 987 to R)
-So only 58.8% is suitable for resistance analysis! We can now filter on it with the filter() function, also from the dplyr package:
+# => Found 2,948 first isolates (59.0% of total)
+So only 59% is suitable for resistance analysis! We can now filter on it with the filter() function, also from the dplyr package:
For future use, the above two syntaxes can be shortened with the filter_first_isolate() function:
@@ -501,63 +501,85 @@
| 1 |
-2010-04-03 |
-C3 |
+2010-05-12 |
+B4 |
B_ESCHR_COL |
-S |
-S |
+R |
+I |
S |
S |
TRUE |
| 2 |
-2010-10-31 |
-C3 |
+2010-05-12 |
+B4 |
B_ESCHR_COL |
R |
S |
-S |
-S |
+R |
+R |
FALSE |
| 3 |
-2010-11-12 |
-C3 |
+2010-07-18 |
+B4 |
B_ESCHR_COL |
+R |
S |
-S |
-S |
+R |
S |
FALSE |
| 4 |
-2010-11-21 |
-C3 |
+2011-01-18 |
+B4 |
B_ESCHR_COL |
R |
-R |
-R |
+S |
+S |
S |
FALSE |
| 5 |
-2010-12-01 |
-C3 |
+2011-11-20 |
+B4 |
B_ESCHR_COL |
S |
+R |
+S |
+S |
+TRUE |
+
+
+| 6 |
+2011-12-03 |
+B4 |
+B_ESCHR_COL |
+S |
+S |
+S |
+S |
+FALSE |
+
+
+| 7 |
+2012-09-09 |
+B4 |
+B_ESCHR_COL |
+R |
S |
S |
S |
FALSE |
-| 6 |
-2011-10-22 |
-C3 |
+8 |
+2013-01-24 |
+B4 |
B_ESCHR_COL |
S |
I |
@@ -566,52 +588,30 @@
TRUE |
-| 7 |
-2012-03-22 |
-C3 |
-B_ESCHR_COL |
-S |
-S |
-S |
-S |
-FALSE |
-
-
-| 8 |
-2012-05-14 |
-C3 |
-B_ESCHR_COL |
-R |
-R |
-S |
-S |
-FALSE |
-
-
| 9 |
-2012-10-26 |
-C3 |
+2013-04-25 |
+B4 |
B_ESCHR_COL |
S |
S |
S |
S |
-TRUE |
+FALSE |
| 10 |
-2013-06-13 |
-C3 |
+2014-02-01 |
+B4 |
B_ESCHR_COL |
S |
+S |
R |
S |
-S |
-FALSE |
+TRUE |
-Only 3 isolates are marked as ‘first’ according to CLSI guideline. But when reviewing the antibiogram, it is obvious that some isolates are absolutely different strains and should be included too. This is why we weigh isolates, based on their antibiogram. The key_antibiotics() function adds a vector with 18 key antibiotics: 6 broad spectrum ones, 6 small spectrum for Gram negatives and 6 small spectrum for Gram positives. These can be defined by the user.
+Only 4 isolates are marked as ‘first’ according to CLSI guideline. But when reviewing the antibiogram, it is obvious that some isolates are absolutely different strains and should be included too. This is why we weigh isolates, based on their antibiogram. The key_antibiotics() function adds a vector with 18 key antibiotics: 6 broad spectrum ones, 6 small spectrum for Gram negatives and 6 small spectrum for Gram positives. These can be defined by the user.
If a column exists with a name like ‘key(…)ab’ the first_isolate() function will automatically use it and determine the first weighted isolates. Mind the NOTEs in below output:
+# => Found 4,405 first weighted isolates (88.1% of total)
| 1 |
-2010-04-03 |
-C3 |
+2010-05-12 |
+B4 |
B_ESCHR_COL |
-S |
-S |
+R |
+I |
S |
S |
TRUE |
@@ -651,46 +651,70 @@
| 2 |
-2010-10-31 |
-C3 |
+2010-05-12 |
+B4 |
B_ESCHR_COL |
R |
S |
-S |
-S |
+R |
+R |
FALSE |
TRUE |
| 3 |
-2010-11-12 |
-C3 |
+2010-07-18 |
+B4 |
B_ESCHR_COL |
+R |
S |
-S |
-S |
+R |
S |
FALSE |
TRUE |
| 4 |
-2010-11-21 |
-C3 |
+2011-01-18 |
+B4 |
B_ESCHR_COL |
R |
-R |
-R |
+S |
+S |
S |
FALSE |
TRUE |
| 5 |
-2010-12-01 |
-C3 |
+2011-11-20 |
+B4 |
B_ESCHR_COL |
S |
+R |
+S |
+S |
+TRUE |
+TRUE |
+
+
+| 6 |
+2011-12-03 |
+B4 |
+B_ESCHR_COL |
+S |
+S |
+S |
+S |
+FALSE |
+TRUE |
+
+
+| 7 |
+2012-09-09 |
+B4 |
+B_ESCHR_COL |
+R |
S |
S |
S |
@@ -698,9 +722,9 @@
TRUE |
-| 6 |
-2011-10-22 |
-C3 |
+8 |
+2013-01-24 |
+B4 |
B_ESCHR_COL |
S |
I |
@@ -710,60 +734,36 @@
TRUE |
-| 7 |
-2012-03-22 |
-C3 |
-B_ESCHR_COL |
-S |
-S |
-S |
-S |
-FALSE |
-FALSE |
-
-
-| 8 |
-2012-05-14 |
-C3 |
-B_ESCHR_COL |
-R |
-R |
-S |
-S |
-FALSE |
-TRUE |
-
-
| 9 |
-2012-10-26 |
-C3 |
+2013-04-25 |
+B4 |
B_ESCHR_COL |
S |
S |
S |
S |
-TRUE |
-TRUE |
+FALSE |
+FALSE |
| 10 |
-2013-06-13 |
-C3 |
+2014-02-01 |
+B4 |
B_ESCHR_COL |
S |
+S |
R |
S |
-S |
-FALSE |
+TRUE |
TRUE |
-Instead of 3, now 9 isolates are flagged. In total, 87.7% of all isolates are marked ‘first weighted’ - 29% more than when using the CLSI guideline. In real life, this novel algorithm will yield 5-10% more isolates than the classic CLSI guideline.
+Instead of 4, now 9 isolates are flagged. In total, 88.1% of all isolates are marked ‘first weighted’ - 29.1% more than when using the CLSI guideline. In real life, this novel algorithm will yield 5-10% more isolates than the classic CLSI guideline.
As with filter_first_isolate(), there’s a shortcut for this new algorithm too:
-So we end up with 4,387 isolates for analysis.
+So we end up with 4,405 isolates for analysis.
We can remove unneeded columns:
@@ -788,28 +788,12 @@
-| 1 |
-2010-10-06 |
-F7 |
-Hospital D |
-B_KLBSL_PNE |
-R |
-S |
-S |
-S |
-M |
-Gram negative |
-Klebsiella |
-pneumoniae |
-TRUE |
-
-
| 2 |
-2015-07-29 |
-T1 |
+2012-11-18 |
+X7 |
Hospital A |
B_ESCHR_COL |
-R |
+S |
S |
S |
S |
@@ -819,68 +803,84 @@
coli |
TRUE |
-
+
| 3 |
-2017-10-20 |
-P2 |
+2012-04-27 |
+K2 |
Hospital B |
-B_STPHY_AUR |
-S |
+B_ESCHR_COL |
+R |
S |
R |
S |
-F |
+M |
+Gram negative |
+Escherichia |
+coli |
+TRUE |
+
+
+| 4 |
+2014-09-03 |
+M6 |
+Hospital A |
+B_STRPTC_PNE |
+S |
+S |
+S |
+R |
+M |
Gram positive |
-Staphylococcus |
-aureus |
+Streptococcus |
+pneumoniae |
TRUE |
-| 4 |
-2010-02-07 |
-Z6 |
+5 |
+2015-09-13 |
+R4 |
Hospital A |
-B_KLBSL_PNE |
-R |
-R |
+B_ESCHR_COL |
+S |
+S |
S |
S |
F |
Gram negative |
-Klebsiella |
-pneumoniae |
+Escherichia |
+coli |
TRUE |
| 6 |
-2016-02-08 |
-S2 |
+2011-09-19 |
+D6 |
Hospital B |
-B_ESCHR_COL |
+B_STRPTC_PNE |
S |
+I |
S |
-S |
-S |
-F |
-Gram negative |
-Escherichia |
-coli |
+R |
+M |
+Gram positive |
+Streptococcus |
+pneumoniae |
TRUE |
-| 9 |
-2016-10-31 |
-H3 |
-Hospital C |
-B_STPHY_AUR |
+7 |
+2013-01-13 |
+Q3 |
+Hospital A |
+B_STRPTC_PNE |
R |
S |
-R |
S |
-M |
+R |
+F |
Gram positive |
-Staphylococcus |
-aureus |
+Streptococcus |
+pneumoniae |
TRUE |
@@ -900,9 +900,9 @@
Or can be used like the dplyr way, which is easier readable:
-Frequency table of genus and species from a data.frame (4,387 x 13)
+
Frequency table of genus and species from a data.frame (4,405 x 13)
Columns: 2
-Length: 4,387 (of which NA: 0 = 0.00%)
+Length: 4,405 (of which NA: 0 = 0.00%)
Unique: 4
Shortest: 16
Longest: 24
@@ -919,33 +919,33 @@ Longest: 24
| 1 |
Escherichia coli |
-2,129 |
-48.5% |
-2,129 |
-48.5% |
+2,160 |
+49.0% |
+2,160 |
+49.0% |
| 2 |
Staphylococcus aureus |
-1,098 |
-25.0% |
-3,227 |
-73.6% |
+1,109 |
+25.2% |
+3,269 |
+74.2% |
| 3 |
Streptococcus pneumoniae |
-688 |
+690 |
15.7% |
-3,915 |
-89.2% |
+3,959 |
+89.9% |
| 4 |
Klebsiella pneumoniae |
-472 |
-10.8% |
-4,387 |
+446 |
+10.1% |
+4,405 |
100.0% |
@@ -956,7 +956,7 @@ Longest: 24
Resistance percentages
The functions portion_R, portion_RI, portion_I, portion_IS and portion_S can be used to determine the portion of a specific antimicrobial outcome. They can be used on their own:
+# [1] 0.4719637
Or can be used in conjuction with group_by() and summarise(), both from the dplyr package:
data_1st %>%
group_by(hospital) %>%
@@ -969,19 +969,19 @@ Longest: 24
| Hospital A |
-0.4544765 |
+0.4711316 |
| Hospital B |
-0.4920107 |
+0.4910714 |
| Hospital C |
-0.4686567 |
+0.4597701 |
| Hospital D |
-0.4570792 |
+0.4488698 |
@@ -999,23 +999,23 @@ Longest: 24
| Hospital A |
-0.4544765 |
-1318 |
+0.4711316 |
+1299 |
| Hospital B |
-0.4920107 |
-1502 |
+0.4910714 |
+1568 |
| Hospital C |
-0.4686567 |
-670 |
+0.4597701 |
+609 |
| Hospital D |
-0.4570792 |
-897 |
+0.4488698 |
+929 |
@@ -1035,27 +1035,27 @@ Longest: 24
| Escherichia |
-0.7491780 |
-0.9074683 |
-0.9798027 |
+0.7407407 |
+0.9000000 |
+0.9763889 |
| Klebsiella |
-0.7521186 |
-0.9067797 |
-0.9745763 |
+0.7533632 |
+0.9147982 |
+0.9843049 |
| Staphylococcus |
-0.7349727 |
-0.9171220 |
-0.9708561 |
+0.7303877 |
+0.9305681 |
+0.9819657 |
| Streptococcus |
-0.7398256 |
+0.7173913 |
0.0000000 |
-0.7398256 |
+0.7173913 |
diff --git a/docs/articles/AMR_files/figure-html/plot 1-1.png b/docs/articles/AMR_files/figure-html/plot 1-1.png
index a326ecf1..ba111e0e 100644
Binary files a/docs/articles/AMR_files/figure-html/plot 1-1.png and b/docs/articles/AMR_files/figure-html/plot 1-1.png differ
diff --git a/docs/articles/AMR_files/figure-html/plot 3-1.png b/docs/articles/AMR_files/figure-html/plot 3-1.png
index 25e8e77b..4302fee6 100644
Binary files a/docs/articles/AMR_files/figure-html/plot 3-1.png and b/docs/articles/AMR_files/figure-html/plot 3-1.png differ
diff --git a/docs/articles/AMR_files/figure-html/plot 4-1.png b/docs/articles/AMR_files/figure-html/plot 4-1.png
index c14b3797..f6efe027 100644
Binary files a/docs/articles/AMR_files/figure-html/plot 4-1.png and b/docs/articles/AMR_files/figure-html/plot 4-1.png differ
diff --git a/docs/articles/AMR_files/figure-html/plot 5-1.png b/docs/articles/AMR_files/figure-html/plot 5-1.png
index bae14fb0..61634b97 100644
Binary files a/docs/articles/AMR_files/figure-html/plot 5-1.png and b/docs/articles/AMR_files/figure-html/plot 5-1.png differ
diff --git a/docs/articles/EUCAST.html b/docs/articles/EUCAST.html
index b56a8d02..7cede9f4 100644
--- a/docs/articles/EUCAST.html
+++ b/docs/articles/EUCAST.html
@@ -40,7 +40,7 @@
AMR (for R)
- 0.5.0.9016
+ 0.5.0.9017
How to apply EUCAST rules
Matthijs S. Berends
- 08 February 2019
+ 09 February 2019
EUCAST.Rmd
diff --git a/docs/articles/G_test.html b/docs/articles/G_test.html
index 7d982d3d..7e04b8c9 100644
--- a/docs/articles/G_test.html
+++ b/docs/articles/G_test.html
@@ -40,7 +40,7 @@
AMR (for R)
- 0.5.0.9016
+ 0.5.0.9017
@@ -185,7 +185,7 @@
How to use the G-test
Matthijs S. Berends
- 08 February 2019
+ 09 February 2019
G_test.Rmd
diff --git a/docs/articles/Predict.html b/docs/articles/Predict.html
index f667f96d..8e0dfce6 100644
--- a/docs/articles/Predict.html
+++ b/docs/articles/Predict.html
@@ -40,7 +40,7 @@
AMR (for R)
- 0.5.0.9016
+ 0.5.0.9017
@@ -185,7 +185,7 @@
How to predict antimicrobial resistance
Matthijs S. Berends
- 08 February 2019
+ 09 February 2019
Predict.Rmd
@@ -194,10 +194,166 @@
-(will be available soon)
+
+
+Needed R packages
+
As with many uses in R, we need some additional packages for AMR analysis. Our package works closely together with the tidyverse packages dplyr and ggplot2 by Dr Hadley Wickham. The tidyverse tremendously improves the way we conduct data science - it allows for a very natural way of writing syntaxes and creating beautiful plots in R.
+
Our AMR package depends on these packages and even extends their use and functions.
+
+
+
+Prediction analysis
+
Our package contains a function resistance_predict(), which takes the same input as functions for other AMR analysis. Based on a date column, it calculates cases per year and uses a regression model to predict antimicrobial resistance.
+
It is basically as easy as:
+
+
# NOTE: Using column `date` as input for `col_date`.
+#
+# Logistic regression model (logit) with binomial distribution
+# ------------------------------------------------------------
+#
+# Call:
+# glm(formula = df_matrix ~ year, family = binomial)
+#
+# Deviance Residuals:
+# Min 1Q Median 3Q Max
+# -2.9224 -1.3120 0.0170 0.7586 3.1932
+#
+# Coefficients:
+# Estimate Std. Error z value Pr(>|z|)
+# (Intercept) -222.92857 45.93922 -4.853 1.22e-06 ***
+# year 0.10994 0.02284 4.814 1.48e-06 ***
+# ---
+# Signif. codes: 0 '***' 0.001 '**' 0.01 '*' 0.05 '.' 0.1 ' ' 1
+#
+# (Dispersion parameter for binomial family taken to be 1)
+#
+# Null deviance: 59.794 on 14 degrees of freedom
+# Residual deviance: 35.191 on 13 degrees of freedom
+# AIC: 93.464
+#
+# Number of Fisher Scoring iterations: 4
+
The function will look for a data column itself if col_date is not set. The result is nothing more than a data.frame, containing the years, number of observations, actual observed resistance, the estimated resistance and the standard error below and above the estimation:
+
+
The function plot is available in base R, and can be extended by other packages to depend the output based on the type of input. We extended its function to cope with resistance predictions:
+
+
+
We also support the ggplot2 package with the function ggplot_rsi_predict():
+
+
+
+
+Choosing the right model
+
Resistance is not easily predicted; if we look at vancomycin resistance in Gram positives, the spread (i.e. standard error) is enormous:
+
+
+
Vancomycin resistance could be 100% in ten years, but might also stay around 0%.
+
You can define the model with the model parameter. The default model is a generalised linear regression model using a binomial distribution, assuming that a period of zero resistance was followed by a period of increasing resistance leading slowly to more and more resistance.
+
Valid values are:
+
+
+
+
+
+
+
+
+
+
+"binomial" or "binom" or "logit"
+ |
+glm(..., family = binomial) |
+Generalised linear model with binomial distribution |
+
+
+
+"loglin" or "poisson"
+ |
+glm(..., family = poisson) |
+Generalised linear model with poisson distribution |
+
+
+
+"lin" or "linear"
+ |
+lm() |
+Linear model |
+
+
+
+
For the vancomycin resistance in Gram positive bacteria, a linear model might be more appropriate since no (left half of a) binomial distribution is to be expected based on observed years:
+
+
+
This seems more likely, doesn’t it?
+
+
How to work with WHONET data
Matthijs S. Berends
- 08 February 2019
+ 09 February 2019
WHONET.Rmd
@@ -227,10 +227,10 @@
Frequency table of mo from a data.frame (500 x 54)
Class: mo (character)
Length: 500 (of which NA: 0 = 0.00%)
-Unique: 56
-Families: 14
-Genera: 23
-Species: 51
+Unique: 39
+Families: 9
+Genera: 16
+Species: 36
| 1 |
B_ESCHR_COL |
-127 |
-25.4% |
-127 |
-25.4% |
+245 |
+49.0% |
+245 |
+49.0% |
| 2 |
B_STPHY_CNS |
-80 |
-16.0% |
-207 |
-41.4% |
+74 |
+14.8% |
+319 |
+63.8% |
| 3 |
-B_STPHY_AUR |
-50 |
-10.0% |
-257 |
-51.4% |
+B_STPHY_EPI |
+38 |
+7.6% |
+357 |
+71.4% |
| 4 |
-B_STPHY_EPI |
-37 |
-7.4% |
-294 |
-58.8% |
-
-
-| 5 |
B_STRPTC_PNE |
31 |
6.2% |
-325 |
-65.0% |
-
-
-| 6 |
-B_STPHY_HOM |
-23 |
-4.6% |
-348 |
-69.6% |
-
-
-| 7 |
-B_PROTS_MIR |
-13 |
-2.6% |
-361 |
-72.2% |
-
-
-| 8 |
-B_KLBSL_PNE |
-11 |
-2.2% |
-372 |
-74.4% |
-
-
-| 9 |
-B_PDMNS_AER |
-8 |
-1.6% |
-380 |
-76.0% |
-
-
-| 10 |
-B_STPHY_CAP |
-8 |
-1.6% |
388 |
77.6% |
+
+| 5 |
+B_STPHY_HOM |
+21 |
+4.2% |
+409 |
+81.8% |
+
+
+| 6 |
+B_PROTS_MIR |
+9 |
+1.8% |
+418 |
+83.6% |
+
+
+| 7 |
+B_ENTRC_IUM |
+8 |
+1.6% |
+426 |
+85.2% |
+
+
+| 8 |
+B_STPHY_CAP |
+8 |
+1.6% |
+434 |
+86.8% |
+
+
+| 9 |
+B_ENTRB_CLO |
+5 |
+1.0% |
+439 |
+87.8% |
+
+
+| 10 |
+B_ENTRC |
+4 |
+0.8% |
+443 |
+88.6% |
+
-(omitted 46 entries, n = 112 [22.4%])
+(omitted 29 entries, n = 57 [11.4%])
# our transformed antibiotic columns
# amoxicillin/clavulanic acid (J01CR02) as an example
@@ -331,9 +331,9 @@ Species: 51
Frequency table of AMC_ND2 from a data.frame (500 x 54)
Class: factor > ordered > rsi (numeric)
Levels: S < I < R
-Length: 500 (of which NA: 41 = 8.20%)
+Length: 500 (of which NA: 19 = 3.80%)
Unique: 3
-%IR: 28.98% (ratio S : IR = 1.0 : 0.4)
+%IR: 25.99% (ratio S : IR = 1.0 : 0.4)
| 1 |
S |
-326 |
-71.0% |
-326 |
-71.0% |
+356 |
+74.0% |
+356 |
+74.0% |
| 2 |
R |
-111 |
-24.2% |
-437 |
-95.2% |
+103 |
+21.4% |
+459 |
+95.4% |
| 3 |
I |
22 |
-4.8% |
-459 |
+4.6% |
+481 |
100.0% |
diff --git a/docs/articles/ab_property.html b/docs/articles/ab_property.html
index 8afb1d42..b0ab510d 100644
--- a/docs/articles/ab_property.html
+++ b/docs/articles/ab_property.html
@@ -40,7 +40,7 @@
AMR (for R)
- 0.5.0.9016
+ 0.5.0.9017
@@ -185,7 +185,7 @@
How to get properties of an antibiotic
Matthijs S. Berends
- 08 February 2019
+ 09 February 2019
ab_property.Rmd
diff --git a/docs/articles/benchmarks.html b/docs/articles/benchmarks.html
index d73a18fe..8d613930 100644
--- a/docs/articles/benchmarks.html
+++ b/docs/articles/benchmarks.html
@@ -40,7 +40,7 @@
AMR (for R)
- 0.5.0.9016
+ 0.5.0.9017
@@ -185,7 +185,7 @@
Benchmarks
Matthijs S. Berends
- 08 February 2019
+ 09 February 2019
benchmarks.Rmd
diff --git a/docs/articles/freq.html b/docs/articles/freq.html
index 5bf39646..45f79d3f 100644
--- a/docs/articles/freq.html
+++ b/docs/articles/freq.html
@@ -40,7 +40,7 @@
AMR (for R)
- 0.5.0.9016
+ 0.5.0.9017
@@ -185,7 +185,7 @@
How to create frequency tables
Matthijs S. Berends
- 08 February 2019
+ 09 February 2019
freq.Rmd
@@ -595,9 +595,9 @@ Unique: 4
Frequency table of amox from a data.frame (2,000 x 49)
Class: factor > ordered > rsi (numeric)
Levels: S < I < R
-Length: 2,000 (of which NA: 828 = 41.40%)
+Length: 2,000 (of which NA: 771 = 38.55%)
Unique: 3
-%IR: 58.53% (ratio S : IR = 1.0 : 1.4)
+%IR: 55.82% (ratio S : IR = 1.0 : 1.3)
| 2 |
S |
-486 |
-41.5% |
-1,169 |
-99.7% |
+543 |
+44.2% |
+1,226 |
+99.8% |
| 3 |
I |
3 |
-0.3% |
-1,172 |
+0.2% |
+1,229 |
100.0% |
@@ -726,9 +726,9 @@ Median: 31 July 2009 (47.39%)
Frequency table of amox from a data.frame (2,000 x 49)
Class: factor > ordered > rsi (numeric)
Levels: S < I < R
-Length: 2,828 (of which NA: 828 = 29.28%)
+Length: 2,771 (of which NA: 771 = 27.82%)
Unique: 4
-%IR: 34.30% (ratio S : IR = 1.0 : 1.4)
+%IR: 34.30% (ratio S : IR = 1.0 : 1.3)
| 1 |
(NA) |
-828 |
-41.4% |
-828 |
-41.4% |
+771 |
+38.6% |
+771 |
+38.6% |
| 2 |
R |
683 |
34.2% |
-1,511 |
-75.6% |
+1,454 |
+72.7% |
| 3 |
S |
-486 |
-24.3% |
+543 |
+27.2% |
1,997 |
99.9% |
diff --git a/docs/articles/mo_property.html b/docs/articles/mo_property.html
index f4475ca1..c9423a29 100644
--- a/docs/articles/mo_property.html
+++ b/docs/articles/mo_property.html
@@ -40,7 +40,7 @@
AMR (for R)
- 0.5.0.9016
+ 0.5.0.9017
@@ -185,7 +185,7 @@
How to get properties of a microorganism
Matthijs S. Berends
- 08 February 2019
+ 09 February 2019
mo_property.Rmd
diff --git a/docs/cover_r4ds.png b/docs/cover_r4ds.png
new file mode 100644
index 00000000..a7150bdf
Binary files /dev/null and b/docs/cover_r4ds.png differ
diff --git a/docs/extra.css b/docs/extra.css
index 362efd1e..7847e8ee 100644
--- a/docs/extra.css
+++ b/docs/extra.css
@@ -21,6 +21,16 @@
# ==================================================================== #
*/
+/* R for Data Science (r4ds) */
+#r4ds a {
+ display: inline;
+}
+#r4ds * {
+ text-align: center;
+ display: block;
+ margin: 0 auto;
+}
+
/* class for footer */
.footer_logo {
float: right;
diff --git a/docs/extra.js b/docs/extra.js
index dce09295..dc4be477 100644
--- a/docs/extra.js
+++ b/docs/extra.js
@@ -34,6 +34,16 @@ $( document ).ready(function() {
window.location.replace(url_new);
}
+ $('#sidebar').prepend(
+ '');
+
$('footer').html(
'' +
'
' + $('footer .copyright p').html().replace(
diff --git a/docs/index.html b/docs/index.html
index 64f43193..a931a8d1 100644
--- a/docs/index.html
+++ b/docs/index.html
@@ -195,7 +195,7 @@
This package can be used for:
- Calculating antimicrobial resistance
-- Predicting empiric susceptibility of both mono therapy and combination therapy
+- Calculating empiric susceptibility of both mono therapy and combination therapy
- Predicting future antimicrobial resistance using regression models
- Getting properties for any microorganism (like Gram stain, species, genus or family)
- Getting properties for any antibiotic (like name, ATC code, defined daily dose or trade name)
diff --git a/docs/news/index.html b/docs/news/index.html
index c1b5e142..d91d6d45 100644
--- a/docs/news/index.html
+++ b/docs/news/index.html
@@ -342,6 +342,7 @@ These functions use as.atc()
- Small text updates to summaries of class
rsi and mic
+- Function
as.rsi() now gives a warning when inputting MIC values
- Frequency tables (
freq() function):
-
diff --git a/docs/reference/AMR-deprecated.html b/docs/reference/AMR-deprecated.html
index 9954eca1..fe2e50a2 100644
--- a/docs/reference/AMR-deprecated.html
+++ b/docs/reference/AMR-deprecated.html
@@ -80,7 +80,7 @@
AMR (for R)
- 0.5.0.9016
+ 0.5.0.9017
' +
'
' + $('footer .copyright p').html().replace(
diff --git a/pkgdown/logos/cover_r4ds.png b/pkgdown/logos/cover_r4ds.png
new file mode 100644
index 00000000..a7150bdf
Binary files /dev/null and b/pkgdown/logos/cover_r4ds.png differ
diff --git a/vignettes/Predict.Rmd b/vignettes/Predict.Rmd
index 56cdbac3..511e8410 100755
--- a/vignettes/Predict.Rmd
+++ b/vignettes/Predict.Rmd
@@ -16,8 +16,96 @@ editor_options:
```{r setup, include = FALSE, results = 'markup'}
knitr::opts_chunk$set(
collapse = TRUE,
- comment = "#"
+ comment = "#",
+ fig.width = 7.5,
+ fig.height = 4.5
)
```
-*(will be available soon)*
+## Needed R packages
+As with many uses in R, we need some additional packages for AMR analysis. Our package works closely together with the [tidyverse packages](https://www.tidyverse.org) [`dplyr`](https://dplyr.tidyverse.org/) and [`ggplot2`](https://ggplot2.tidyverse.org) by [Dr Hadley Wickham](https://www.linkedin.com/in/hadleywickham/). The tidyverse tremendously improves the way we conduct data science - it allows for a very natural way of writing syntaxes and creating beautiful plots in R.
+
+Our `AMR` package depends on these packages and even extends their use and functions.
+
+```{r lib packages, message = FALSE}
+library(dplyr)
+library(ggplot2)
+library(AMR)
+
+# (if not yet installed, install with:)
+# install.packages(c("tidyverse", "AMR"))
+```
+
+## Prediction analysis
+Our package contains a function `resistance_predict()`, which takes the same input as functions for [other AMR analysis](./articles/AMR.html). Based on a date column, it calculates cases per year and uses a regression model to predict antimicrobial resistance.
+
+It is basically as easy as:
+```{r, eval = FALSE}
+# resistance prediction of piperacillin/tazobactam (pita):
+resistance_predict(tbl = septic_patients, col_date = "date", col_ab = "pita")
+
+# or:
+septic_patients %>%
+ resistance_predict(col_ab = "pita")
+
+# to bind it to object 'predict_pita' for example:
+predict_pita <- septic_patients %>%
+ resistance_predict(col_ab = "pita")
+```
+
+```{r, echo = FALSE}
+predict_pita <- septic_patients %>%
+ resistance_predict(col_ab = "pita")
+```
+
+The function will look for a data column itself if `col_date` is not set. The result is nothing more than a `data.frame`, containing the years, number of observations, actual observed resistance, the estimated resistance and the standard error below and above the estimation:
+```{r}
+predict_pita
+```
+
+The function `plot` is available in base R, and can be extended by other packages to depend the output based on the type of input. We extended its function to cope with resistance predictions:
+
+```{r}
+plot(predict_pita)
+```
+
+We also support the `ggplot2` package with the function `ggplot_rsi_predict()`:
+
+```{r}
+library(ggplot2)
+ggplot_rsi_predict(predict_pita)
+```
+
+### Choosing the right model
+
+Resistance is not easily predicted; if we look at vancomycin resistance in Gram positives, the spread (i.e. standard error) is enormous:
+
+```{r}
+septic_patients %>%
+ filter(mo_gramstain(mo) == "Gram positive") %>%
+ resistance_predict(col_ab = "vanc", year_min = 2010, info = FALSE) %>%
+ plot()
+```
+
+Vancomycin resistance could be 100% in ten years, but might also stay around 0%.
+
+You can define the model with the `model` parameter. The default model is a generalised linear regression model using a binomial distribution, assuming that a period of zero resistance was followed by a period of increasing resistance leading slowly to more and more resistance.
+
+Valid values are:
+
+| Input values | Function used by R | Type of model |
+|----------------------------------------|-------------------------------|-----------------------------------------------------|
+| `"binomial"` or `"binom"` or `"logit"` | `glm(..., family = binomial)` | Generalised linear model with binomial distribution |
+| `"loglin"` or `"poisson"` | `glm(..., family = poisson)` | Generalised linear model with poisson distribution |
+| `"lin"` or `"linear"` | `lm()` | Linear model |
+
+For the vancomycin resistance in Gram positive bacteria, a linear model might be more appropriate since no (left half of a) binomial distribution is to be expected based on observed years:
+
+```{r}
+septic_patients %>%
+ filter(mo_gramstain(mo) == "Gram positive") %>%
+ resistance_predict(col_ab = "vanc", year_min = 2010, info = FALSE, model = "linear") %>%
+ plot()
+```
+
+This seems more likely, doesn't it?
' +
+ '