How to dynamically create a number of columns and insert an element into each

Using Streamlit
1

My app would involve allowing the user to choose the number of inputs he wishes to enter
For example, at the top, there would be a select box asking the user for the number of inputs, say 1 to 5, I’ll call this n_input.

And then, I would like there to be that number of select boxes to show up. So for example, the user selected “3” in the previous n_input select box, then 3 select boxes would show up which allow the user to input 3 items.

This is not that difficult to accomplish:
for i in range(n_input):
input[i]=st.selectbox("Input # "+str(i),[1,2,3],key=i)

However, I don’t know how to pack them into columns instead of just lining up vertically. The number of variables to unpack ranges from 1 to 7 so this requires some dynamic assignment.

columns[1],columns[2],.....columns[n_words]=st.beta_columns(nwords)

Surely, I could make a big if loop,

if n_input==7:
     columns[1],columns[2],...,columns[7]=st.beta_columns(7)
if n_input==6:
     columns[1],columns[2],...,columns[6]=st.beta_columns(6)
if n_input==5:
     columns[1],columns[2],...,columns[5]=st.beta_columns(5)
....

Is there a better way to do this?

2

Hi @terbiume65, welcome to the Streamlit community!

You can accomplish what you are talking about using the following:

import streamlit as st


ncol = st.sidebar.number_input("Number of dynamic columns", 0, 20, 1)
cols = st.beta_columns(ncol)

for i, x in enumerate(cols):
    x.selectbox(f"Input # {i}",[1,2,3], key=i)

Best,
Randy

3 Likes
3

Hi, just to follow up, is there any way to set a max number of columns (e.g. 3) to show per row? Say if I have three items to show, they’ll be on the same row, but if there’re 5 items, they’ll be split into two rows, 7 items 3 rows, etc?
Thanks!

4

Hi Hurley,

You can use the Python’s modulo operator to do the wrapping. The modulo operator (%) returns the remainder of 2 numbers e.g. 5%3 returns 2 which would put the fifth element in the second column of the second row.

import streamlit as st

ncol = st.sidebar.number_input("Number of columns", 0, 20, 1)
wcol = st.sidebar.number_input("Wrap at column", 1, 3, 1)

cols = st.columns(ncol)

for i in range(ncol):
    col = cols[i%wcol]
    col.selectbox(f"Input # {i}",[1,2,3], key=i)
5 Likes
5

Hi @hudgeon,

Is there a way to accomplish this without the selectbox, but with graphs instead.

For example:

ncol = count_trues #(which is the number trues in a dataframe column)
wcol = 5

cols = st.columns(ncol)

for i in range(ncol):
      col = cols[i%wcol]
      with col:
          #my graphs

Thanks!