Draft new data_summary() function

[strengejacke]

Equvalent to dplyr::summarise()

library(datawizard)
data(iris)
data_summary(iris, MW = mean(Sepal.Width), SD = sd(Sepal.Width))
#>   MW |   SD
#> -----------
#> 3.06 | 0.44

data_summary(
  iris,
  MW = mean(Sepal.Width),
  SD = sd(Sepal.Width),
  by = "Species"
)
#> Species    |   MW |   SD
#> ------------------------
#> setosa     | 3.43 | 0.38
#> versicolor | 2.77 | 0.31
#> virginica  | 2.97 | 0.32

# same as
d <- data_group(iris, "Species")
data_summary(d, MW = mean(Sepal.Width), SD = sd(Sepal.Width))
#> Species    |   MW |   SD
#> ------------------------
#> setosa     | 3.43 | 0.38
#> versicolor | 2.77 | 0.31
#> virginica  | 2.97 | 0.32

# multiple groups
data(mtcars)
data_summary(mtcars, MW = mean(mpg), SD = sd(mpg), by = c("am", "gear"))
#> am | gear |    MW |   SD
#> ------------------------
#>  0 |    3 | 16.11 | 3.37
#>  0 |    4 | 21.05 | 3.07
#>  1 |    4 | 26.27 | 5.41
#>  1 |    5 | 21.38 | 6.66

# expressions can also be supplied as character strings
data_summary(mtcars, "MW = mean(mpg)", "SD = sd(mpg)", by = c("am", "gear"))
#> am | gear |    MW |   SD
#> ------------------------
#>  0 |    3 | 16.11 | 3.37
#>  0 |    4 | 21.05 | 3.07
#>  1 |    4 | 26.27 | 5.41
#>  1 |    5 | 21.38 | 6.66

^{Created on 2024-03-03 with reprex v2.1.0}

[codecov]

Codecov Report

Attention: Patch coverage is 84.17722% with 25 lines in your changes are missing coverage. Please review.

Project coverage is 90.47%. Comparing base (83f9703) to head (af73c75).

Files	Patch %	Lines
R/categorize.R	56.75%	16 Missing ⚠️
R/adjust.R	53.84%	6 Missing ⚠️
R/data_summary.R	95.83%	3 Missing ⚠️

Additional details and impacted files

@@            Coverage Diff             @@
##             main     #482      +/-   ##
==========================================
+ Coverage   90.34%   90.47%   +0.13%     
==========================================
  Files          73       74       +1     
  Lines        5696     5776      +80     
==========================================
+ Hits         5146     5226      +80     
  Misses        550      550              

☔ View full report in Codecov by Sentry.
📢 Have feedback on the report? Share it here.

[strengejacke]

I moved some of the code from data_modify() in a separate function, to avoid duplicated code.

[strengejacke]

Some examples with NA values:

library(datawizard)
data(efc, package = "datawizard")

data_summary(efc, MW = mean(c12hour, na.rm = TRUE), by = "c172code")
#> c172code |    MW
#> ----------------
#>        1 | 87.12
#>        2 | 94.05
#>        3 | 75.00
#>     <NA> | 47.80

data_summary(efc, MW = mean(c12hour, na.rm = TRUE), by = "c172code", include_na = FALSE)
#> c172code |    MW
#> ----------------
#>        1 | 87.12
#>        2 | 94.05
#>        3 | 75.00

data_summary(efc, MW = mean(c12hour, na.rm = TRUE), by = c("e42dep", "c172code"))
#> e42dep | c172code |     MW
#> --------------------------
#> 1      |        2 |  17.00
#> 2      |        2 |  34.25
#> 3      |        1 |  39.50
#> 3      |        2 |  52.44
#> 3      |        3 |  52.00
#> 3      |     <NA> |  84.00
#> 4      |        1 | 134.75
#> 4      |        2 | 119.26
#> 4      |        3 |  88.80
#> 4      |     <NA> |  43.29
#> <NA>   |        2 |   <NA>
#> <NA>   |     <NA> |   7.00

^{Created on 2024-03-03 with reprex v2.1.0}

[DominiqueMakowski]

Would be nice to make it compatible with our dataframe-returning functions, like ci() etc.

[strengejacke]

Would be nice to make it compatible with our dataframe-returning functions, like ci() etc.

See #483

I guess this would require special handling for these "exceptions" (separate handling for each special class). But let's first finish this PR.

[strengejacke]

@etiennebacher if you want to review this PR, please wait until all checks pass.

[etiennebacher]

Yes I'll review it in a couple of hours

[strengejacke]

This reduced coverage is due to my lintr-fix in adjust.R, actually the code-coverage increased in this PR.

[etiennebacher]

Thanks @strengejacke, LGTM but I have a few comments