Time-Series Lag Feature

Build a one-step lag feature from an ordered, non-empty numeric series: position $tt$ holds the value from position $t-1t-1$, and position $00$ has no predecessor so it is NaN.

Implement solve(x).

Returns. Return a float NumPy array of the same length as the input.

Examples

Example 1

Input:

solve([10.0, 20.0, 30.0, 40.0])

Output:

[nan, 10.0, 20.0, 30.0]

Example 2

Input:

solve([7.0])

Output:

[nan]