Text and Number Formatting

These formatting functions are used to format numerical values in a consistent manner. This is useful for printing numbers inline with text, as well as for formatting tables. Many of the included formatting functions were adapted from TJ Mahr's printy package.

fmt_count(x, big_interval = 3L, big_mark = ",")

fmt_digits(x, digits = 3)

fmt_leading_zero(x)

fmt_minus(x, output = NULL)

fmt_replace_na(x, replacement = "&mdash;")

fmt_corr(x, digits, output = NULL)

fmt_prop(x, digits, fmt_small = TRUE, keep_zero = FALSE)

fmt_prop_pct(x, digits = 0, fmt_small = TRUE)

Arguments

x: Number or number string to be formatted
big_interval: Interval indicating where to place numeric dividers
big_mark: Character used as mark between big interval before the decimal
digits: Number of decimal places to retain
output: The output type for the rendered document. One of "latex" or "html".
replacement: The value to use when replacing missing values
fmt_small: Indicator for replacing zero with < (e.g., .000 becomes <.001). Default is TRUE.
keep_zero: If fmt_small is TRUE, whether to preserve true 0s (e.g., 0.0000001 becomes <.001, but 0.0000000 stays .000).

Details

fmt_digits() is a wrapper for base::sprintf(). Prints a number with digits number of decimal places, without losing trailing zeros, as happens with base::round().

fmt_leading_zero() removes the leading zero for decimal values.

fmt_minus() replaces hyphens with the HTML minus sign (−).

fmt_replace_na() replaces NA values with a specified replacement. This is useful for formatting tables, when blanks are not desired. The default behavior is to replace missing values with an em-dash (—).

fmt_prop_pct() formats proportions as percentages. This takes a number bounded between 0 and 1, multiplies it by 100, and then rounds to the specified number of digits using fmt_digits().

Two additional formatters are provided to format numbers according to the American Psychological Association (APA) style guide. The 7th edition of the APA style guide specifies that numbers bounded between [-1, 1] should not include the leading zero (section 6.36; APA, 2020). This is the case for many types of numbers commonly used by ATLAS including correlations, proportions, probabilities, and p-values. The fmt_corr() function is used to format values bounded between [-1, 1]. Digits are first rounded to the specified number of digits using fmt_digits(), and then leading zeros are removed using fmt_leading_zero() and negative signs are replaced with fmt_minus(). The fmt_prop is very similar, but is intended for values between [0, 1]. This function also wraps fmt_digits() and fmt_leading_zero(). However, fmt_prop() also replaces small values to avoid values of 0 (e.g., .00 is replaced with < .01).

References

American Psychological Association. (2020). Publication manual of the American Psychological Association (7th ed.). https://doi.org/10.1037/0000165-000

Examples

test_cor <- cor(mtcars[, 1:4])
as.character(round(test_cor[1:4, 3], 2))
#> [1] "-0.85" "0.9"   "1"     "0.79" 
fmt_digits(test_cor[1:4, 3], 2)
#> [1] "-0.85" "0.90"  "1.00"  "0.79" 

fmt_digits(test_cor[1:4, 3], 2) %>%
  fmt_leading_zero()
#> [1] "-.85" ".90"  "1.00" ".79" 

fmt_digits(test_cor[1:4, 3], 2) %>%
  fmt_minus()
#> [1] "&minus;0.85" "0.90"        "1.00"        "0.79"       

fmt_digits(c(test_cor[1:4, 3], NA_real_), 2) %>%
  fmt_replace_na(replacement = "&mdash;")
#> [1] "-0.85"   "0.90"    "1.00"    "0.79"    "&mdash;"

fmt_corr(test_cor[1:4, 3], 2)
#> [1] "&minus;.85" ".90"        "1.00"       ".79"       

fmt_prop(c(0.001, 0.035, 0.683), digits = 2)
#> [1] "<.01" ".04"  ".68"

Arguments

Details

References

See also

Examples