# Select colums by the value of rows
**URL:** https://discourse.julialang.org/t/select-colums-by-the-value-of-rows/84793
**Category:** General Usage
**Tags:** question, dataframes
**Created:** [July 26, 2022, 6:23am UTC](https://discourse.julialang.org/t/select-colums-by-the-value-of-rows/84793 "2022-07-26T06:23:02Z")
**Posts on this page:** 5
**Page:** 1
### Author:  [@chenr86](https://discourse.julialang.org/u/chenr86)
#### Post date: [July 26, 2022, 6:23am UTC](https://discourse.julialang.org/t/select-colums-by-the-value-of-rows/84793/1 "2022-07-26T06:23:03Z")
Hello,
I want to select target colums with row values between an interval. For example, with df like below, I want to select colums with row values between [3, 5], which means that it should return `:x1 :x2 :y1`. How can I do that?
```julia
df = DataFrame(a=1, b=2, x1=3, x2=4, y1=5, y2=6)
```
---
### Author:  [@chenr86](https://discourse.julialang.org/u/chenr86)
#### Post date: [July 26, 2022, 7:49am UTC](https://discourse.julialang.org/t/select-colums-by-the-value-of-rows/84793/2 "2022-07-26T07:49:07Z")
For someone interested on this topic, the solution can be:
```julia
df_filter = DataFrame()
for i in 1:size(df, 2)
if 3 <= df[end, i] <= 5
df_filter[!, "Plan$i"] = df[:, i]
end
end
df_filter
```
---
### Author:  [@Jollywatt](https://discourse.julialang.org/u/Jollywatt)
#### Post date: [July 26, 2022, 8:08am UTC](https://discourse.julialang.org/t/select-colums-by-the-value-of-rows/84793/3 "2022-07-26T08:08:47Z")
You could also select columns via a predicate like so:
```julia
julia> collect(df[1, :])
6-element Vector{Int64}:
1
2
3
4
5
6
julia> 3 .<= ans .<= 5
6-element BitVector:
0
0
1
1
1
0
julia> df[!, ans] # use ! to return a view; use : to copy
1×3 DataFrame
Row │ x1 x2 y1
│ Int64 Int64 Int64
─────┼─────────────────────
1 │ 3 4 5
```
---
### Author:  [@DataFrames](https://discourse.julialang.org/u/DataFrames)
#### Post date: [August 5, 2022, 4:29am UTC](https://discourse.julialang.org/t/select-colums-by-the-value-of-rows/84793/4 "2022-08-05T04:29:09Z")
use
```julia
df[:, [all(in(3:5), x) for x in eachcol(df)]]
```
---
### Author:  [@rafael.guerra](https://discourse.julialang.org/u/rafael.guerra)
#### Post date: [August 5, 2022, 2:32pm UTC](https://discourse.julialang.org/t/select-colums-by-the-value-of-rows/84793/5 "2022-08-05T14:32:55Z")
Perhaps safer to not assume integers:
```julia
df[:, [all(3 .≤ x .≤ 5) for x in eachcol(df)]]
```