Query for seeing dates on which integers in another column change

[Nash]

Suppose I have a Dataframe. One of the columns contains Dates sorted from low to high. Another columns contains integers. On particular dates, the value in the integer-column changes. What query will allow me to find see the dates on which the value in the integer-column has changed?

[jling]

I want to say just use a for-loop but you can also do this I suppose

julia> df = DataFrame(dates=today()-Day(5):Day(1):today(), ints = [1,1,2,3,3,4])
6×2 DataFrame
 Row │ dates       ints  
     │ Date        Int64 
─────┼───────────────────
   1 │ 2021-10-19      1
   2 │ 2021-10-20      1
   3 │ 2021-10-21      2
   4 │ 2021-10-22      3
   5 │ 2021-10-23      3
   6 │ 2021-10-24      4

julia> df[[0;diff(df.ints)] .!= 0, :]
3×2 DataFrame
 Row │ dates       ints  
     │ Date        Int64 
─────┼───────────────────
   1 │ 2021-10-21      2
   2 │ 2021-10-22      3
   3 │ 2021-10-24      4

[rafael.guerra]

If one also wants to keep the first date as a reference, an alternative:

df.dates[df.ints .!= circshift(df.ints,1)]

4-element Vector{Date}:
 2021-10-19
 2021-10-21
 2021-10-22
 2021-10-24