Yesterday's Average

A fitted imputer fills every later gap with the number it learned from the training rows, never with an average of the new rows.

Task: write fill_later(train_values, later_values, strategy).

• Both lists hold numbers with some None gaps. train_values has at least one real number.
• Fit a SimpleImputer(strategy=strategy) on train_values only. strategy is "mean" or "median".
• Use it to fill the gaps in later_values.
• Return [learned value, filled later values]. The learned value is the imputer's statistics_ entry. Round it and every filled value to 2 decimal places.