JuliaForDataAnalysis/errata.md

Errata

This file contains errata for the "Julia for Data Analysis" book that has been written by Bogumił Kamiński and has been published by Manning Publications Co.

Chapter 1, section 1.2.1, page 7

I show the following example of code execution:

julia> function sum_n(n)
           s = 0
           for i in 1:n
               s += i
           end
           return s
       end
sum_n (generic function with 1 method)

julia> @time sum_n(1_000_000_000)
  0.000001 seconds
500000000500000000

This timing is very fast (and the reason is explained in the book). The issue is that this is the situation under Julia 1.7.

Under Julia 1.8 and Julia 1.9 running the same code takes longer (tested under Julia 1.9-beta4):

julia> @time sum_n(1_000_000_000)
  2.265569 seconds
500000000500000000

The reason for this inconsistency is a bug in the @time macro introduced in Julia 1.8 release. The sum_n(1_000_000_000) call (without @time) is executed fast. Here is a simplified benchmark (run under Julia 1.9-beta4):

julia> let
           start = time_ns()
           v = sum_n(1_000_000_000)
           stop=time_ns()
           v, Int(stop - start)
       end
(500000000500000000, 1000)

Unfortunately there is an issue with the @time macro used in global scope, that needs to be resolved in Base Julia. See this issue.

Chapter 2, introduction, page 20

• middle of page 20: the provided link http://mng.bz/5mWD explaining k-times winsorized mean definition no longer works. Use https://web.archive.org/web/20210804184830/https://v8doc.sas.com/sashtml/insight/chap38/sect17.htm provided by The Wayback Machine instead.

Chapter 2, section 2.3.1, page 30

I compare the following expressions:

x > 0 && println(x)

and

if x > 0
    println(x)
end

where x = -7.

I write there that Julia interprets them both in the same way. It is true in terms of the fact that in both cases the println function is not called (and this is the focus point of the example). However, there is a difference in the value of these expressions. The first expression evaluates to false, while the second evaluates to nothing.

Here is how you can check it:

julia> x = -7
-7

julia> show(x > 0 && println(x))
false
julia> show(if x > 0
           println(x)
       end)
nothing

Chapter 2, section 2.5, page 45

• top of page 45: use in this book): should be use in this book:

Chapter 3, section 3.2.3, page 58

• middle of page 58: y[end - the + 1] = y[end -- k] should be y[end - i + 1] = y[end - k]

Chapter 3, section 3.2.3, page 59

• top of page 59: sort(v::AbstractVector; kwthe.) should be sort(v::AbstractVector; kws...)

Chapter 6, section 6.4.1, page 132

• middle of Listing 6.4: codeunits("?") should be codeunits("ε")

Chapter 8, section 8.1.2, page 189

• middle of page 189: zsdf format should be zstd format

Chapter 8, section 8.2.1, page 191

• bottom of page 191: misssingstring should be missingstring

Chapter 9, section 9.2.2, page 231

• top of page 191: both ratings should be ratings

Chapter 10, section 10.2.2, page 255

• bottom of page 255: ? Error: Error adding value to column :b. should be ┌ Error: Error adding value to column :b.

Chapter 12, section 12.1.4, page 302

• bottom of page 302:

julia> df = DataFrame(a=1:3, b=1:3, c=1:3)
3×3 DataFrame
 Row │ a      b      c
     │ Int64  Int64  Int64
???????????????????????????
   1 │     1      1      1
   2 │     2      2      2
   3 │     3      3      3

should be

julia> df = DataFrame(a=1:3, b=1:3, c=1:3)
3×3 DataFrame
 Row │ a      b      c
     │ Int64  Int64  Int64
─────┼─────────────────────
   1 │     1      1      1
   2 │     2      2      2
   3 │     3      3      3

Chapter 12, section 12.3.2, page 318

• top of page 318: in the annotation to Figure 12.6 there is text Applies a log1p which looks like Applies a loglp (this is a display issue due to the fact that in the font used letter l and digit 1 look identical)