Effectively, there are many functions for carrying out anova

If the repetitions involve different units, you are in a "classical" analysis of variance. You can use the lm function.

If repetitions are made on the same units, account must be taken of the possible source of variation due to responses that are similar for the same individual. Different methods are possible:

* If you are interested only in averages (comparison of treatments), you can use marginal models such as generalized least squares (package nlme, gls function) or GEE models (geese function of the geepack package)

* If you are interested in the effects of individuals, you can also use mixed effect models such as those adjusted by the lme function of the nlme package, or lmer the Matrix and lme4 packages.

These models allow to analyze unbalanced experimental plans.

An important question is whether the missing data occur independently of the process studied:

* If not, the results are biased,

* If this is the case, attention must be paid to the interpretation of the estimated parameters (conditional on the missing data).

For a general overview of methods suitable for repeated data, see for example:

Diggle PJ, Liang K-Y and Zeger SL, 1994. Analysis of longitudinal data, 1st ed. Clarendon Press, Oxford.

or

https://cran.r-project.org/doc/contrib/Faraway-PRA.pdf

thanks

This may be helpful:

http://rcompanion.org/handbook/I_09.html

Also, somewhat simpler, see the Rattlesnake example at the end of this page, using either nlme or lmer for mixed effects model.

http://rcompanion.org/rcompanion/d_08.html