Introduction to helpr

The helpr package contains numerous helpers, wrappers, and utilities used throughout the my analysis suite. It intentionally favors base R over higher level tidyverse to minimize imports. The goal is to provide an alternative gain in functionality without the cost of additional imports/dependencies.

---

Useful functions in helpr

• File-system: See ?filesystem
• Signal UI: See ?signal
• Symbols: See ?sybml
• Strings:
  • squish()
  • trim()
  • pad()
  • capture()
• Handlers:
  • be_safe()
  • be_quiet()
  • be_hard()
• liter():
  • List iteration, similar to purrr::imap() but simpler
• Calculations:
  • calc_ccc(): Calculate Lin’s Concordance Correlation Coefficient for two vectors of numeric data.
  • calc_ss(): Calculate the sum of squared errors: $\sum{(x - \bar x)^2}$ for numeric data.
• Logic Tests:
  • rep_lgl(): Are all the elements of a vector identical?
  • is_monotonic(): Are the numerical elements of a vector monotonically increasing or decreasing?
  • is_logspace(): Does a given object (vector, data.frame, tbl_df) appear to be in log-space? Note: this function is biased to proteomic data and should not be expected to be accurate for other applications out-of-the-box.
• cross_tab()
  • Create a contingency table of counts generated by cross-classifying.
• diff_vecs()
  • Generate all diffs of two vectors of (typically character) data. Returns both sided setdiff, union and intersect.
• dater()
  • Generate a standardized data in (by default) YYYY-MM-DD.

---

Examples

cross_tab()

You do not need to “quote” the passed arguments, un-quoted strings are fine and are parsed by NSE (non-standard evaluation). The ... can take on either one or two column names:

# 1 factor
cross_tab(mtcars, cyl)      # unquoted string
#> cyl
#>   4   6   8 Sum 
#>  11   7  14  32

cross_tab(mtcars, "cyl")    # quoted string
#> cyl
#>   4   6   8 Sum 
#>  11   7  14  32

var <- "cyl"
cross_tab(mtcars, var)      # external variable
#> cyl
#>   4   6   8 Sum 
#>  11   7  14  32

cross_tab(mtcars, cyl, gear)      # 2 factors
#>      gear
#> cyl    3  4  5 Sum
#>   4    1  8  2  11
#>   6    2  4  1   7
#>   8   12  0  2  14
#>   Sum 15 12  5  32

cross_tab(mtcars, cyl, gear, am)  # 3 factors
#> , , am = 0
#> 
#>      gear
#> cyl    3  4  5 Sum
#>   4    1  2  0   3
#>   6    2  2  0   4
#>   8   12  0  0  12
#>   Sum 15  4  0  19
#> 
#> , , am = 1
#> 
#>      gear
#> cyl    3  4  5 Sum
#>   4    0  6  2   8
#>   6    0  2  1   3
#>   8    0  0  2   2
#>   Sum  0  8  5  13
#> 
#> , , am = Sum
#> 
#>      gear
#> cyl    3  4  5 Sum
#>   4    1  8  2  11
#>   6    2  4  1   7
#>   8   12  0  2  14
#>   Sum 15 12  5  32

calc_ccc()

Calculate the ccc for two numeric vectors (visualize by concordance):

x <- rnorm(100, mean = 10, sd = 0.5)
y <- x + rnorm(100, sd = 0.1)   # add random scatter
ccc <- calc_ccc(x, y)
plot(x, y, pch = 21, cex = 1.75, col = NA,
     bg = rgb(red = 0, green = 0, blue = 0, alpha = 0.5),   # black w alpha
     main = sprintf("The CCC = %0.3f with significance of p = %0.3f",
                    ccc$rho.c, ccc$p.value)
)
abline(0, 1, lty = 2, col = "blue", lwd = 1.5)

Concordance plot visualizing calc_ccc().

ccc
#> $rho.c
#> [1] 0.9778699
#> 
#> $ci95
#>     lower     upper 
#> 0.9675816 0.9849182 
#> 
#> $Z
#> [1] 2.24642
#> 
#> $p.value
#> [1] 0.02467715

calc_ss()

Calculate the sum of squared differences for a numeric vector. Used ubiquotously in generating variances and standard deviations within other contexts (e.g. ANOVA, CVs):

x <- rnorm(100, mean = 10, sd = 0.5)
calc_ss(x)
#> [1] 18.85493

calc_ss(x) / (length(x) - 1)    # variance
#> [1] 0.1904538

all.equal(var(x), calc_ss(x) / (length(x) - 1))    # TRUE
#> [1] TRUE

rep_lgl()

rep_lgl(letters)
#> [1] FALSE

rep_lgl(rep("A", 250))
#> [1] TRUE

rep_lgl(c("B", rep("A", 250)))
#> [1] FALSE

is_monotonic()

is_monotonic(1:100)
#> [1] TRUE

is_monotonic(seq(-100, 100, by = 5))     # up
#> [1] TRUE

is_monotonic(seq(100, -100, by = -5))    # down
#> [1] TRUE

is_monotonic(rnorm(10))
#> [1] FALSE

is_logspace()

# A numeric vector
x <- rnorm(30, mean = 1000)
is_logspace(x)
#> [1] FALSE

is_logspace(log(x))
#> [1] TRUE

is_logspace(data) # FALSE

# log10-transform
for ( i in grep("^ft", names(data)) ) {
  data[[i]] <- log10(data[[i]])
}

is_logspace(data) # base 10; TRUE

diff_vecs()

diff_vecs(LETTERS[1:10L], LETTERS[5:15L], verbose = TRUE) # return invisible
#> ℹ Vectors differ by:• Unique to LETTERS[1:10L] >> 4• Unique to
#> LETTERS[5:15L] >> 5• Common Intersect >> 6• Union >> 15

(diff_vecs(LETTERS[1:10L], LETTERS[5:15L]))
#> $`unique_LETTERS[1:10L]`
#> [1] "A" "B" "C" "D"
#> 
#> $`unique_LETTERS[5:15L]`
#> [1] "K" "L" "M" "N" "O"
#> 
#> $inter
#> [1] "E" "F" "G" "H" "I" "J"
#> 
#> $unique
#>  [1] "A" "B" "C" "D" "E" "F" "G" "H" "I" "J" "K" "L" "M" "N" "O"

dater()

data <- data.frame(x = rnorm(100), y = rnorm(100))
plot(data$x, data$y, main = paste0("This is today's date: ", dater()))

Standardized date format using dater().