This function creates lagged choice covariates from the `data.frame`

`choice_data`

, which is assumed to be sorted by the choice occasions.

## Arguments

- choice_data
A

`data.frame`

of choice data in wide format, i.e. each row represents one choice occasion.- column
A character, the column name in

`choice_data`

, i.e. the covariate name. Can be a vector.- k
A positive number, the number of lags (in units of observations), see the details. Can be a vector. The default is

`k = 1`

.- id
A character, the name of the column in

`choice_data`

that contains unique identifier for each decision maker. The default is`"id"`

.

## Value

The input `choice_data`

with the additional columns named
`column.k`

for each element `column`

and each number `k`

containing the lagged covariates.

## Details

Say that `choice_data`

contains the column `column`

. Then, the
function call

```
create_lagged_cov(choice_data, column, k, id)
```

returns the input `choice_data`

which includes a new column named
`column.k`

. This column contains for each decider (based on `id`

)
and each choice occasion the covariate faced before `k`

choice
occasions. If this data point is not available, it is set to
`NA`

. In particular, the first `k`

values of `column.k`

will
be `NA`

(initial condition problem).