RCC SU 2018

.Rbuildignore

@@ -9,3 +9,4 @@
 ^pkgdown$
 ^cran-comments\.md$
 ^CRAN-SUBMISSION$
+^data-raw$

NEWS.md

@@ -1,5 +1,7 @@
 # surveytable (development version)
 
+* `rccsu2018`
+
 # surveytable 0.9.4
 
 * Optionally adjust p-values for multiple comparisons (`p_adjust` argument)

R/z_data.R

@@ -26,3 +26,13 @@
 #' United States as of July 1, 2019. Used for calculating rates. For usage
 #' examples, see the `*_rate` functions.
 "uspop2019"
+
+#' National Study of Long-Term Care Providers (NSLTCP) Residential Care Community (RCC) Services User (SU) 2018 Public Use File (PUF)
+#'
+#' A data system of RCC residents.
+#'
+#' @source
+#' * SAS data: <https://ftp.cdc.gov/pub/Health_Statistics/NCHS/Datasets/NPALS/final2018rcc_su_puf.sas7bdat>
+#' * Documentation: <https://www.cdc.gov/nchs/npals/RCCresident-readme03152021vr.pdf>
+#' * Codebook: <https://www.cdc.gov/nchs/data/npals/final2018rcc_su_puf_codebook.pdf>
+"rccsu2018"

README.Rmd

@@ -18,7 +18,26 @@ knitr::opts_chunk$set(
 <!-- badges: start -->
 <!-- badges: end -->
 
-The `surveytable` package provides short and understandable commands that generate tabulated, formatted, and rounded survey estimates. One useful function, which operates on categorical and logical variables, tabulates estimated counts and percentages with their standard errors and confidence intervals. Other functions estimate the total population, tabulate survey subsets and variable interactions, tabulate numeric variables, perform hypothesis tests, tabulate rates, modify survey variables, and save the output. All of the tabulation functions identify low-precision estimates using algorithms developed at the National Center for Health Statistics (NCHS). If the `surveytable` code is called from an R Markdown notebook, it automatically generates HTML or LaTeX tables, as appropriate. The package reduces the number of commands that users need to execute, which is especially helpful for users new to R or to programming.
+`surveytable` is an R package for conveniently tabulating estimates from **complex surveys**.
+
+* If you deal with survey objects in R (created with `survey::svydesign()`), then this package is for you.
+* Works with **complex surveys** (data systems that involve survey design variables, like weights and strata).
+* Works with **unweighted** data as well.
+
+* The `surveytable` package provides short and understandable commands that generate tabulated, formatted, and rounded survey estimates. 
+* With `surveytable`, you can
+  * tabulate estimated counts and percentages, with their standard errors and confidence intervals,
+  * estimate the total population, 
+  * tabulate survey subsets and variable interactions, 
+  * tabulate numeric variables, 
+  * perform hypothesis tests, 
+  * tabulate rates, 
+  * modify survey variables, and 
+  * save the output. 
+
+* All of the tabulation functions identify low-precision estimates using National Center for Health Statistics (NCHS) algorithms. 
+* If the `surveytable` code is called from an R Markdown notebook or a Quarto document, it automatically generates HTML or LaTeX tables, as appropriate. 
+* The package reduces the number of commands that users need to execute, which is especially helpful for users new to R or to programming.
 
 
 ## Installation
@@ -28,7 +47,7 @@ Install from CRAN:
 ``` r
 install.packages("surveytable")
 ```
-or get a development version from GitHub:
+or get the development version from GitHub:
 
 ``` r
 install.packages(c("remotes", "git2r"))
@@ -39,20 +58,39 @@ remotes::install_github("CDCgov/surveytable", upgrade = "never")
 
 Find the documentation for `surveytable` here: https://cdcgov.github.io/surveytable/
 
-
 ## Example
 
 Here is a basic example, to get you started. 
 
-```{r, results=FALSE, message=FALSE, warning=FALSE}
+1. Load the package:
+
+```{r, message=FALSE}
 library(surveytable)
 ```
 
+2. Specify the survey that you wish you analyze. `surveytable` comes with a survey called `namcs2019sv`, for use in examples. 
+
 ```{r, results='asis'}
 set_survey(namcs2019sv)
+```
+
+3. Specify the variable to analyze. In NAMCS, `AGER` is the age category variable:
+
+```{r, results='asis'}
 tab("AGER")
 ```
 
+The table shows:
+
+* Descriptive variable name
+* Survey name
+* For each level of the variable:
+  * Number of observations
+  * Estimated count with its SE and 95% CI
+  * Estimated percentage with its SE and 95% CI
+* Sample size
+* Whether any low-precision estimates were found
+
 <!-- CDC standard text -->
 ## Public Domain Standard Notice
 This repository constitutes a work of the United States Government and is not

README.md

@@ -6,20 +6,41 @@
 <!-- badges: start -->
 <!-- badges: end -->
 
-The `surveytable` package provides short and understandable commands
-that generate tabulated, formatted, and rounded survey estimates. One
-useful function, which operates on categorical and logical variables,
-tabulates estimated counts and percentages with their standard errors
-and confidence intervals. Other functions estimate the total population,
-tabulate survey subsets and variable interactions, tabulate numeric
-variables, perform hypothesis tests, tabulate rates, modify survey
-variables, and save the output. All of the tabulation functions identify
-low-precision estimates using algorithms developed at the National
-Center for Health Statistics (NCHS). If the `surveytable` code is called
-from an R Markdown notebook, it automatically generates HTML or LaTeX
-tables, as appropriate. The package reduces the number of commands that
-users need to execute, which is especially helpful for users new to R or
-to programming.
+`surveytable` is an R package for conveniently tabulating estimates from
+**complex surveys**.
+
+- If you deal with survey objects in R (created with
+  `survey::svydesign()`), then this package is for you.
+
+- Works with **complex surveys** (data systems that involve survey
+  design variables, like weights and strata).
+
+- Works with **unweighted** data as well.
+
+- The `surveytable` package provides short and understandable commands
+  that generate tabulated, formatted, and rounded survey estimates.
+
+- With `surveytable`, you can
+
+  - tabulate estimated counts and percentages, with their standard
+    errors and confidence intervals,
+  - estimate the total population,
+  - tabulate survey subsets and variable interactions,
+  - tabulate numeric variables,
+  - perform hypothesis tests,
+  - tabulate rates,
+  - modify survey variables, and
+  - save the output.
+
+- All of the tabulation functions identify low-precision estimates using
+  National Center for Health Statistics (NCHS) algorithms.
+
+- If the `surveytable` code is called from an R Markdown notebook or a
+  Quarto document, it automatically generates HTML or LaTeX tables, as
+  appropriate.
+
+- The package reduces the number of commands that users need to execute,
+  which is especially helpful for users new to R or to programming.
 
 ## Installation
 
@@ -29,7 +50,7 @@ Install from CRAN:
 install.packages("surveytable")
 ```
 
-or get a development version from GitHub:
+or get the development version from GitHub:
 
 ``` r
 install.packages(c("remotes", "git2r"))
@@ -45,17 +66,22 @@ Find the documentation for `surveytable` here:
 
 Here is a basic example, to get you started.
 
+1.  Load the package:
+
 ``` r
 library(surveytable)
 ```
 
+2.  Specify the survey that you wish you analyze. `surveytable` comes
+    with a survey called `namcs2019sv`, for use in examples.
+
 ``` r
 set_survey(namcs2019sv)
 ```
 
 <table class="huxtable" data-quarto-disable-processing="true" style="border-collapse: collapse; border: 0px; margin-bottom: 2em; margin-top: 2em; ; margin-left: auto; margin-right: auto;  ">
 <caption style="caption-side: top; text-align: center;">
-Survey info {NAMCS 2019 PUF}
+Survey info {namcs2019sv}
 </caption>
 <col>
 <col>
@@ -80,19 +106,22 @@ Design
 </td>
 <td style="vertical-align: top; text-align: left; white-space: normal; border-style: solid solid solid solid; border-width: 0.4pt 0.4pt 0.4pt 0.4pt;    padding: 6pt 6pt 6pt 6pt; font-weight: normal;">
 Stratified 1 - level Cluster Sampling design (with replacement)<br>With
-(398) clusters.<br>survey::svydesign(ids = ~CPSUM, strata = ~CSTRATM,
-weights = ~PATWT, <br> data = namcs2019sv_df)
+(398) clusters.<br>namcs2019sv = survey::svydesign(ids = ~CPSUM, strata
+= ~CSTRATM, weights = ~PATWT, data = namcs2019sv_df)
 </td>
 </tr>
 </table>
 
+3.  Specify the variable to analyze. In NAMCS, `AGER` is the age
+    category variable:
+
 ``` r
 tab("AGER")
 ```
 
 <table class="huxtable" data-quarto-disable-processing="true" style="border-collapse: collapse; border: 0px; margin-bottom: 2em; margin-top: 2em; ; margin-left: auto; margin-right: auto;  ">
 <caption style="caption-side: top; text-align: center;">
-Patient age recode {NAMCS 2019 PUF}
+Patient age recode {namcs2019sv}
 </caption>
 <col>
 <col>
@@ -334,6 +363,18 @@ N = 8250. Checked NCHS presentation standards. Nothing to report.
 </td>
 </tr>
 </table>
+
+The table shows:
+
+- Descriptive variable name
+- Survey name
+- For each level of the variable:
+  - Number of observations
+  - Estimated count with its SE and 95% CI
+  - Estimated percentage with its SE and 95% CI
+- Sample size
+- Whether any low-precision estimates were found
+
 <!-- CDC standard text -->
 
 ## Public Domain Standard Notice

_pkgdown.yml

@@ -31,6 +31,7 @@ reference:
   contents:
   - namcs2019sv
   - uspop2019
+  - rccsu2018
 - title: "Other"
   contents:
   - set_output

data-raw/namcs2019sv.R

@@ -0,0 +1,29 @@
+library(nchsdata)
+library(magrittr)
+
+vrs = c("CPSUM", "CSTRATM", "PATWT" # survey design
+        , "MDDO", "SPECCAT", "MSA"  # Table 1
+        , "AGER", "SEX", "AGE"      # Table 3
+        , "NOPAY", "PAYPRIV", "PAYMCARE", "PAYMCAID", "PAYWKCMP"
+        , "PAYOTH", "PAYDK", "PAYSELF", "PAYNOCHG" # Table 5
+        , "PRIMCARE", "REFER", "SENBEFOR" # Table 6
+        , "MAJOR" # Table 11
+        , "NUMMED" # numeric variable
+        , "ANYIMAGE", "BONEDENS", "CATSCAN", "ECHOCARD", "OTHULTRA"
+        , "MAMMO", "MRI", "XRAY", "OTHIMAGE" # imaging
+)  %>% unique
+
+namcs2019sv_df = namcs2019$variables[,vrs]
+
+vr = "SPECCAT.bad"
+namcs2019sv_df[,vr] = namcs2019sv_df$SPECCAT
+set.seed(42)
+nr = nrow(namcs2019sv_df)
+idx = sample.int(n = nr, size = round(0.2 * nr))
+namcs2019sv_df[idx, vr] = NA
+attr(namcs2019sv_df[,vr], "label") = "Type of specialty (BAD - do not use)"
+
+namcs2019sv = survey::svydesign(ids = ~CPSUM, strata = ~CSTRATM, weights = ~PATWT, data = namcs2019sv_df)
+
+usethis::use_data(namcs2019sv, overwrite = TRUE)
+usethis::use_data(namcs2019sv_df, overwrite = TRUE)

data-raw/rccsu2018.R

@@ -0,0 +1,2 @@
+library(nchsdata)
+usethis::use_data(rccsu2018, overwrite = TRUE)