Dear all,

I have a dataframe and I can use describe to have an initial hint of the data. But to have more control, I would like to find these values one column at the time.

So I can find the average of one column with `mean(df[:column])`

.

But if I do `min(df[:column])`

I get the error `MethodError No method matching min`

, but `using statistics`

is loaded.

And if I do `describe(df[:column], stats = :min)`

I get an argument error.

What is wrong?

Can I apply these basic statistics to selected columns of a dataframe? Can I find a median or quartile/percentiles?

Thank you

You want `minimum(df[:column])`

. Note that you can also do `df.column`

. The `min`

function is for returning the minimum of its arguments, for example `min(0,1,2) == 0`

. For `describe`

, the `stats`

keyword argument is deprecated. Use regular arguments instead, for example `describe(df, :mean, :min)`

.

Also note that you can view the documentation string for these functions directly in the REPL by typing e.g. `?describe`

.

If you are not already aware of it, you may want to check out StatsBase.jl.

1 Like

Thank you. I use REPL but I did not know about minimum. I’ll have a look at StatsBase.

1 Like