included fill_na

TidierOrg · Aug 3, 2023 · 4db0700 · 4db0700
1 parent 4e51309
commit 4db0700
Show file tree

Hide file tree

Showing 2 changed files with 47 additions and 2 deletions.
diff --git a/src/TidierData.jl b/src/TidierData.jl
@@ -15,7 +15,7 @@ using Reexport
 @reexport using ShiftedArrays: lag, lead
 
 export TidierData_set, across, desc, n, row_number, starts_with, ends_with, matches, if_else, case_when, ntile, 
-      as_float, as_integer, as_string, @select, @transmute, @rename, @mutate, @summarize, @summarise, @filter,
+      as_float, as_integer, as_string, fill_na, @select, @transmute, @rename, @mutate, @summarize, @summarise, @filter,
       @group_by, @ungroup, @slice, @arrange, @distinct, @pull, @left_join, @right_join, @inner_join, @full_join,
       @pivot_wider, @pivot_longer, @bind_rows, @bind_cols, @clean_names, @count, @tally, @drop_na, @glimpse, @separate,
       @unite, @summary

diff --git a/src/docstrings.jl b/src/docstrings.jl
@@ -2099,4 +2099,49 @@ julia> @chain df begin
    2 │ C          11    12.0      13.0  12.6667    13.5      14      3              2
    3 │ D          16    17.0      18.0  18.0       19.0      20      5              0
 ```
-"""
+"""
+
+const = docstring_fill_na
+"""
+   fill(column, method)
+
+Fills missing values in a given column of a DataFrame. This function uses either Last Observation Carried Forward (LOCF) or Next Observation Carried Backward (NOCB) methods based on the provided argument. 
+
+# Arguments
+
+- `column`: Column that with missing values
+- `method`: "nocb" or "locf" depending on the desired strategy. 
+```jldoctest
+
+julia> df = DataFrame(dt1=[missing, 0.2, missing, missing, 1, missing, 5, 6], dt2=[0.3, missing, missing, 3, missing, 5, 6,missing])
+8×2 DataFrame
+ Row │ dt1        dt2       
+     │ Float64?   Float64?  
+─────┼──────────────────────
+   1 │ missing          0.3
+   2 │       0.2  missing   
+   3 │ missing    missing   
+   4 │ missing          3.0
+   5 │       1.0  missing   
+   6 │ missing          5.0
+   7 │       5.0        6.0
+   8 │       6.0  missing   
+   
+julia> @chain df begin
+       @mutate(dt1 = ~fill_na(dt1, "locf"))
+       @mutate(dt2 = ~fill_na(dt2, "nocb"))
+       end
+8×2 DataFrame
+ Row │ dt1        dt2       
+     │ Float64?   Float64?  
+─────┼──────────────────────
+   1 │ missing          0.3
+   2 │       0.2        3.0
+   3 │       0.2        3.0
+   4 │       0.2        3.0
+   5 │       1.0        5.0
+   6 │       1.0        5.0
+   7 │       5.0        6.0
+   8 │       6.0  missing   
+```
+"""