Merge pull request #96 from TidierOrg/fix-and-or-vectorization

Fixed auto-vectorization of `&&` and `||` and added docs.

NEWS.md

@@ -1,5 +1,9 @@
 # Tidier.jl updates
 
+## v0.7.5 - 2023-04-30
+- Fixed bug to ensure that `&&` and `||` are auto-vectorized
+- Added docstrings and examples to show different ways of filtering by multiple "and" conditions, including `&&`, `&`, and separating multiple expressions with commas.
+
 ## v0.7.4 - 2023-04-11
 - Added `as_float()`, `as_integer()`, and `as_string()`
 

Project.toml

@@ -1,7 +1,7 @@
 name = "Tidier"
 uuid = "f0413319-3358-4bb0-8e7c-0c83523a93bd"
 authors = ["Karandeep Singh"]
-version = "0.7.4"
+version = "0.7.5"
 
 [deps]
 Chain = "8be319e6-bccf-4806-a6f7-6fae938471bc"

docs/examples/UserGuide/filter.jl

@@ -8,10 +8,36 @@ movies = dataset("ggplot2", "movies");
 # Let’s take a look at the movies whose budget was more than average. We will select only the first 5 rows for the sake of brevity.
 
 @chain movies begin
-    @mutate(Budget = Budget / 1_000_000)
-    @filter(Budget >= mean(skipmissing(Budget)))
-    @select(Title, Budget)
-    @slice(1:5)
+  @mutate(Budget = Budget / 1_000_000)
+  @filter(Budget >= mean(skipmissing(Budget)))
+  @select(Title, Budget)
+  @slice(1:5)
+end
+
+# Let's search for movies that have at least 200 votes and a rating of greater than or equal to 8. There are 3 ways you can specify an "and" condition inside of `Tidier.jl`.
+
+## The first option is to use the short-circuiting `&&` operator as shown below. This is the preferred approach because the second expression is only evaluated (per element) if the first one is true.
+
+@chain movies begin
+  @filter(Votes >= 200 && Rating >= 8)
+  @select(Title, Votes, Rating)
+  @slice(1:5)
+end
+
+## The second option is to use the bitwise `&` operator. However, there is a key difference in syntax. Because the `&` operator takes a higher operator precendence than `>=`, you have to wrap the `>=` expressions inside of parentheses to ensure that the expression is evaluated correctly.
+
+@chain movies begin
+  @filter((Votes >= 200) & (Rating >= 8))
+  @select(Title, Votes, Rating)
+  @slice(1:5)
+end
+
+## Finally, for "and" conditions only, you can separate the expressions with commas, similar to the behavior of `filter()` in `tidyverse`.
+
+@chain movies begin
+  @filter(Votes >= 200, Rating >= 8)
+  @select(Title, Votes, Rating)
+  @slice(1:5)
 end
 
 # Now let's see how to use `@filter()` with `in`. Here's an example with a tuple.

src/docstrings.jl

@@ -475,6 +475,16 @@ julia> @chain df begin
    2 │ d         4     14
    3 │ e         5     15
 
+julia> @chain df begin
+       @filter(b >= 3 && c >= 14)
+       end
+2×3 DataFrame
+ Row │ a     b      c     
+     │ Char  Int64  Int64 
+─────┼────────────────────
+   1 │ d         4     14
+   2 │ e         5     15
+
 julia> @chain df begin
        @filter(b in (1, 3))
        end

src/parsing.jl

@@ -302,6 +302,9 @@ function parse_autovec(tidy_expr::Union{Expr,Symbol})
         fn_new = Symbol("." * string(fn))
         return :($fn_new($(args...)))
       end
+    elseif hasproperty(x, :head) && (x.head == :&& || x.head == :||)
+      x.head = Symbol("." * string(x.head))
+      return x
     end
     return x
   end