# `cube_aggregate_time()`

**Category:** Temporal aggregation
**Status:** experimental

## Purpose

Aggregate only historical time into day, ISO-week, calendar-month,
meteorological-season, or calendar-year periods.

## Real signature

```r
cube_aggregate_time(
  x, by, method = "mean", na.rm = TRUE, min_n = 1L,
  diagnostics = FALSE
)
```

Finite stored observations have equal weight. Supported reducers are mean,
sampled-value sum, minimum, maximum, and exact median. Date remains Date;
POSIXct remains UTC POSIXct. Internal empty periods are retained, and optional
diagnostics report observation counts rather than duration coverage.

Lazy NetCDF values are reduced through bounded period and spatial/depth/
variable reads. The full source is not materialized as an intermediate; the
regularized result is an in-memory five-dimensional cube.

## Minimal executable check

```r
library(oceancube)
stopifnot(identical(
  names(formals(cube_aggregate_time)),
  c("x", "by", "method", "na.rm", "min_n", "diagnostics")
))
```

Typical next operations are `cube_climatology()`, `cube_anomaly()`, or
`cube_trend()`. Use `to_month()` only for compatibility.