Helper function that extracts year from a dataframe and raises error if more than one.

MCOD files name the year column `year` from data year 1996 onward but `datayear` for 1979-1995. This checks `year` first, then falls back to `datayear`, and errors if neither column is present.

Usage

.extract_year(df)

Arguments

df

dataframe to extract year from

Value

year as integer (four-digit)

Details

The 1979-1995 files store `datayear` as a two-digit value (e.g. 85 for 1985). A two-digit value is normalized to its four-digit year (+1900), since every two-digit `datayear` is a 19xx year. Without this, downstream year dispatch (`.dispatch_era()`, `remap_race()`, `categorize_hspanicr()`) sees a value like 85 that matches no coding era and either errors or silently mislabels.