rescale() can expand ranges

[strengejacke]

Similar to scales::expand_range(), rescale() can now also expand ranges:

library(datawizard)
x <- seq(5, 15, 2)
x
#> [1]  5  7  9 11 13 15
rescale(x, multiply = 1.1)
#> [1]  4.5  6.7  8.9 11.1 13.3 15.5
#> (original range = 5 to 15)

rescale(x, add = 1)
#> [1]  4.0  6.4  8.8 11.2 13.6 16.0
#> (original range = 5 to 15)

rescale(x, add = -2)
#> [1]  7.0  8.2  9.4 10.6 11.8 13.0
#> (original range = 5 to 15)

rescale(x, add = c(3, 10))
#> [1]  2.0  6.6 11.2 15.8 20.4 25.0
#> (original range = 5 to 15)

rescale(x, add = c(-3, 5))
#> [1]  8.0 10.4 12.8 15.2 17.6 20.0
#> (original range = 5 to 15)


d <- data.frame(x = seq(5, 15, 2), y = seq(5, 15, 2))

# multiply range of each column, expanding by 10%
rescale(d, multiply = 1.1)
#>      x    y
#> 1  4.5  4.5
#> 2  6.7  6.7
#> 3  8.9  8.9
#> 4 11.1 11.1
#> 5 13.3 13.3
#> 6 15.5 15.5

# add 0.5 to each end of range (same as 10% expanding)
rescale(d, add = 0.5)
#>      x    y
#> 1  4.5  4.5
#> 2  6.7  6.7
#> 3  8.9  8.9
#> 4 11.1 11.1
#> 5 13.3 13.3
#> 6 15.5 15.5

# multiply range of x by 1.1 and range of y by 0.5
rescale(d, multiply = list(x = 1.1, y = 0.5))
#>      x    y
#> 1  4.5  7.5
#> 2  6.7  8.5
#> 3  8.9  9.5
#> 4 11.1 10.5
#> 5 13.3 11.5
#> 6 15.5 12.5

# add 0.5 to each end of range of x and subtract 2.5 from each end of range of y
rescale(d, add = list(x = 0.5, y = -2.5))
#>      x    y
#> 1  4.5  7.5
#> 2  6.7  8.5
#> 3  8.9  9.5
#> 4 11.1 10.5
#> 5 13.3 11.5
#> 6 15.5 12.5

# add 1 to lower end of range of x and 3 to upper end of range of x
# add 2 to lower end of range of y and 4 to upper end of range of y
rescale(d, add = list(x = c(1, 3), y = c(2, 4)))
#>      x    y
#> 1  4.0  3.0
#> 2  6.8  6.2
#> 3  9.6  9.4
#> 4 12.4 12.6
#> 5 15.2 15.8
#> 6 18.0 19.0

Not sure, though, if add should add (subtract) the value to each end of the range? I.e. new range = c(min - add, max + add)? Then the range would expand by 2*add.

[codecov]

Codecov Report

All modified and coverable lines are covered by tests ✅

Comparison is base (b2d3b1b) 89.03% compared to head (a81ae2d) 89.09%.

❗ Current head a81ae2d differs from pull request most recent head a6e73b8. Consider uploading reports for the commit a6e73b8 to get more accurate results

Additional details and impacted files

@@            Coverage Diff             @@
##             main     #467      +/-   ##
==========================================
+ Coverage   89.03%   89.09%   +0.05%     
==========================================
  Files          72       72              
  Lines        5383     5409      +26     
==========================================
+ Hits         4793     4819      +26     
  Misses        590      590              

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

[strengejacke]

@bwiernik @etiennebacher we could add an alias expand_range(), which only has arguments multiply and add (and not to, like rescale()), which then calls rescale().

[etiennebacher]

Thanks, minor things

[etiennebacher]

Not sure, though, if add should add (subtract) the value to each end of the range?

Hard to say, I don't have a use case for this function in mind, I guess you have one?

we could add an alias expand_range()

wouldn't this create a conflict with scales::expand_range() (which is called a lot by ggplot2 IIRC)?

[strengejacke]

Not sure, though, if add should add (subtract) the value to each end of the range?

Hard to say, I don't have a use case for this function in mind, I guess you have one?

It seems that expected behaviour in this case would be to add add to min/max (I have changed this here already), see https://bsky.app/profile/tjmahr.com/post/3ke3nc3hix32u

we could add an alias expand_range()

wouldn't this create a conflict with scales::expand_range() (which is called a lot by ggplot2 IIRC)?

True. So no alias.

[etiennebacher]

I'm not on bluesky, can you share a screenshot? (sorry I edited your post instead of writing a comment)

[strengejacke]

Oh, I thought it would be visible for non-registered, too:

If you like, I can send you an invitation code.

[etiennebacher]

If you like, I can send you an invitation code.

I'm fine on Mastodon for now, thanks ;)

[strengejacke]

Social networks have become less important anyway, or have pretty much fallen asleep since Musk took over Twitter.

[etiennebacher]

Social networks have become less important anyway, or have pretty much fallen asleep since Musk took over Twitter.

I don't know, I just can't understand the hype for bluesky, I don't see any reason why it will end up differently than twitter, but anyway that's not the topic 😄

[bwiernik]

Personally I like Threads a lot among the various options