# Dividing a Dataframe

**URL:** <https://discourse.julialang.org/t/dividing-a-dataframe/131631>\
**Category:** General Usage\
**Tags:** dataframes\
**Created:** [August 15, 2025, 3:07pm UTC](https://discourse.julialang.org/t/dividing-a-dataframe/131631 "2025-08-15T15:07:25Z")\
**Posts on this page:** 7\
**Page:** 1

<div class="post-metadata">

**Author:** ![SergeantMike67](https://sea2.discourse-cdn.com/julialang/user_avatar/discourse.julialang.org/sergeantmike67/32/25103_2.png) [@SergeantMike67](https://discourse.julialang.org/u/SergeantMike67)\
**Post date:** [August 15, 2025, 3:07pm UTC](https://discourse.julialang.org/t/dividing-a-dataframe/131631/1 "2025-08-15T15:07:25Z")

</div>

I have a dataframe

```julia-auto
 Row │ VarietyA VarietyB Variety.p.adj 
     │ String String Float64
─────┼─────────────────────────────────────────────────────────────────────────────────
   1 │ Backwoods Blueberries Coville 128th Northland 0.0011105
   2 │ Backwoods Blueberries Jersey 128th Northland 0.98911
   3 │ Bazan Duke 128th Northland 0.559359
   4 │ Bazan Jersey 128th Northland 0.746563
   5 │ Brown Site BlueJay 128th Northland 0.00315533
   6 │ Brown Site Jersey 128th Northland 0.0131208
   7 │ Otis Lake V. corymbosum 128th Northland 0.183493
   8 │ Russells Blueberry Farm and Book… 128th Northland 0.0792741
   9 │ Talsma Aurora 128th Northland 0.946847
  10 │ Talsma BlueJay 128th Northland 0.858454
  11 │ Talsma Liberty 128th Northland 0.00410839
  12 │ Wolf Lake V. angustifolium 128th Northland 0.360707
  13 │ Backwoods Blueberries Jersey Backwoods Blueberries Coville 0.0715494
  14 │ Bazan Duke Backwoods Blueberries Coville 0.0858854
  15 │ Bazan Jersey Backwoods Blueberries Coville 0.795704
  16 │ Brown Site BlueJay Backwoods Blueberries Coville 1.27e-7
  17 │ Brown Site Jersey Backwoods Blueberries Coville 5.15e-7
  18 │ Otis Lake V. corymbosum Backwoods Blueberries Coville 3.91e-6
  19 │ Russells Blueberry Farm and Book… Backwoods Blueberries Coville 0.999625
  20 │ Talsma Aurora Backwoods Blueberries Coville 2.34e-5

```

What I would Like to do is find all the instances of Variety X (where X is any one of 13 Varieties) in VarietyA and VarietyB and put them into that into a grouped dataframe. similar to the table below.

| VarietyA | VarietyB | Variety.p.adj |
| --- | --- | --- |
| Brown Site Jersey | 128th Northland | 0.013120833 |
| Brown Site Jersey | Backwoods Blueberries Coville | 5.15E-07 |
| Brown Site Jersey | Backwoods Blueberries Jersey | 0.00564981 |
| Brown Site Jersey | Bazan Duke | 0.000177299 |
| Brown Site Jersey | Bazan Jersey | 0.005307901 |
| Brown Site Jersey | Brown Site BlueJay | 0.999987571 |
| Otis Lake V. corymbosum | Brown Site Jersey | 0.925359246 |
| Russells Blueberry Farm and Book Barn Northland | Brown Site Jersey | 0.000111835 |
| Talsma Aurora | Brown Site Jersey | 0.082719439 |
| Talsma BlueJay | Brown Site Jersey | 0.008174087 |
| Talsma Liberty | Brown Site Jersey | 1.26E-06 |
| Wolf Lake V. angustifolium | Brown Site Jersey | 4.94E-05 |

The idea of this is to put all of one variety together to make it easier to interpret the results.

Thanks  
Mike

---

<div class="post-metadata">

**Author:** ![jling](https://sea2.discourse-cdn.com/julialang/user_avatar/discourse.julialang.org/jling/32/212909_2.png) [@jling](https://discourse.julialang.org/u/jling)\
**Post date:** [August 15, 2025, 3:22pm UTC](https://discourse.julialang.org/t/dividing-a-dataframe/131631/2 "2025-08-15T15:22:28Z")

</div>

> [@SergeantMike67](#):
>
> put them into that into a grouped dataframe. similar to the table below.

but in this table, neither VarietyA or VarietyB is unique? do you want `sort` the dataframe or `group` them?

---

<div class="post-metadata">

**Author:** ![Jeff\_Emanuel](https://sea2.discourse-cdn.com/julialang/user_avatar/discourse.julialang.org/jeff_emanuel/32/15440_2.png) [@Jeff\_Emanuel](https://discourse.julialang.org/u/Jeff_Emanuel)\
**Post date:** [August 15, 2025, 3:28pm UTC](https://discourse.julialang.org/t/dividing-a-dataframe/131631/3 "2025-08-15T15:28:43Z")

</div>

Perhaps `filter([:VarietyA, :VarietyB] => (a, b) -> a == "Brown Site Jersey" || b == "Brown Site Jersey", df)`. Then you could `sort` with a custom `lt` that prefers “Brown Site Jersey” in each column.

You won’t be able to group for each of your 13 varieties, because in a GroupedDataframe, a row can only belong to a single group.

---

<div class="post-metadata">

**Author:** ![jling](https://sea2.discourse-cdn.com/julialang/user_avatar/discourse.julialang.org/jling/32/212909_2.png) [@jling](https://discourse.julialang.org/u/jling)\
**Post date:** [August 15, 2025, 3:35pm UTC](https://discourse.julialang.org/t/dividing-a-dataframe/131631/4 "2025-08-15T15:35:54Z")

</div>

oh it’s a `or` condition, in that case:

```julia-auto
mask = @. df.VarietyA == "Brown Site Jersey" || df.VarietyB == "Brown Site Jersey"
df[mask, :]

```

will do?

---

<div class="post-metadata">

**Author:** ![technocrat](https://sea2.discourse-cdn.com/julialang/user_avatar/discourse.julialang.org/technocrat/32/220947_2.png) [@technocrat](https://discourse.julialang.org/u/technocrat)\
**Post date:** [August 15, 2025, 7:06pm UTC](https://discourse.julialang.org/t/dividing-a-dataframe/131631/5 "2025-08-15T19:06:07Z")

</div>

DataFrames has `subset` for that.

```julia-auto
using DataFrames

df = DataFrame(a = ["apple", "banana", "cherry"], b = ["apple", "orange", "apricot"])
subset(df, [:a, :b] => ByRow((a, b) -> a == b))
1×2 DataFrame
 Row │ a b      
     │ String String 
─────┼────────────────
   1 │ apple apple

```

Read this as “find the variables in df where row a and row b are the same”

---

<div class="post-metadata">

**Author:** ![Jeff\_Emanuel](https://sea2.discourse-cdn.com/julialang/user_avatar/discourse.julialang.org/jeff_emanuel/32/15440_2.png) [@Jeff\_Emanuel](https://discourse.julialang.org/u/Jeff_Emanuel)\
**Post date:** [August 15, 2025, 7:51pm UTC](https://discourse.julialang.org/t/dividing-a-dataframe/131631/6 "2025-08-15T19:51:03Z")

</div>

Either `filter` or `subset` works fine for this case, but `filter` works by row, so it doesn’t require `ByRow` here. On the other hand, `subset` works by column, so it is more general. The DataFrames documentation encourages preference for `subset`:

> This method is defined so that DataFrames.jl implements the Julia API for collections, but it is generally recommended to use the [`subset`](https://dataframes.juliadata.org/stable/lib/functions/#DataFrames.subset) function instead as it is consistent with other DataFrames.jl functions (as opposed to `filter`).

See also [Filter and Subset - Julia Data Science](https://juliadatascience.io/filter_subset)

---

<div class="post-metadata">

**Author:** ![pdeffebach](https://sea2.discourse-cdn.com/julialang/user_avatar/discourse.julialang.org/pdeffebach/32/10320_2.png) [@pdeffebach](https://discourse.julialang.org/u/pdeffebach)\
**Post date:** [August 15, 2025, 9:54pm UTC](https://discourse.julialang.org/t/dividing-a-dataframe/131631/7 "2025-08-15T21:54:54Z")

</div>

with DataFramesMeta.jl you have

```julia-auto
df_sub = @rsubset df :VarietyA == "Brown Site Jersey" || :VarietyB == "Brown Site Jersey" 

```