# Predict method for moult models Predict either the proportion of birds in a certain moult stage (as defined in intervals) on a specified day, the average moult score on a specified day, or start and/or duration of moult for given covariate patterns ```r ## S3 method for class 'moult' predict(object, newdata, predict.type = "prob", intervals = 0.1, ...) ``` ## Arguments - `object`: moult model objects - `newdata`: optional dataframe with explanatory variables for which to make predictions. The first column must contain the days (as used when fitting) for which to make predictions. - `predict.type`: specifies form of predictions, see details. - `intervals`: length of moult categories for which probabilities/proportions should be calculated. The default (= 0.1) will calculate the probability that a bird will fall in this moult category on the specified day for each of the following categories: 0, (0, 0.1), [0.1, 0.2), [0.2, 0.3), ..., [0.9, 1.0), 1. - `...`: other arguments passed to the predict method ## Details `predict.type` has the following options: - **"response"**: the average moult index (proportion of feather mass grown) for each of the days specified. - **"prob"**: default, the proportion of birds in each of the moult categories as defined by intervals is predicted. - **"duration"**: predicts the duration of moult for the covariate combinations defined in `newdata`, with standard errors. If `newdata` is not supplied, returns duration of baseline covariate set. - **"start"**: predicts the mean start date of moult for the covariate combinations defined in `newdata`, with standard errors. If `newdata` is not supplied, returns mean start date of baseline covariate set. - **"both"**: predicts both mean start date of moult and duration of moult for the covariate combinations defined in `newdata`, with standard errors. Also, covariance of duration and start date estimates is given. ## Returns If `newdata` is missing, the expected moult scores at the observed days are returned. If `predict.type = "response"`, the expected moult scores at the specified days are returned. If `predict.type = "prob"` a matrix of predicted probabilities for being in each of the moult categories defined by intervals. If `predict.type` equals "start" or "duration" or "both", the corresponding estimates (with standard errors) for each of the covariate patterns are returned. ## References Erni, B., Bonnevie, B. T., Oschadleus, H.-D., Altwegg, R. and Underhill, L. G. (2013) moult: An R package to analyze moult in birds. Journal of Statistical Software, 52 (8), 1--23. [https://www.jstatsoft.org/v52/i08/](https://www.jstatsoft.org/v52/i08/) Underhill, L. G. and Zucchini, W. (1988) A model for avian primary moult. **Ibis** 130 , 358--372. Underhill, L. G. and Zucchini, W. and Summers, R. W. (1990) A model for avian primary moult-data types based on migration strategies and an example using the Redshank **Tringa totanus**. **Ibis** 132 , 118--123. ## Author(s) birgit.erni@uct.ac.za ## See Also `moult` ## Examples ```r data(weavers) ## convert moult scores to PFMG (proportion feather mass grown) mscores <- substr(weavers$Moult, 1, 9) feather.mass <- c(10.4, 10.8, 11.5, 12.8, 14.4, 15.6, 16.3, 15.7, 15.7) weavers$pfmg <- ms2pfmg(mscores, feather.mass) ## day of year starting 1 August weavers$day <- date2days(weavers$RDate, dateformat = "yyyy-mm-dd", startmonth = 8) weavers$ssex <- ifelse(weavers$Sex == 1 | weavers$Sex == 3, "male", ifelse(weavers$Sex == 2 | weavers$Sex == 4, "female", NA)) mmf <- moult(pfmg ~ day | ssex | ssex, data = weavers, type = 3) summary(mmf) ## predict duration and start of moult (then both) for males and females ssex <- c("male", "female") day <- 150 (p1 <- predict.moult(mmf, newdata = data.frame(day, ssex), predict.type = "duration")) (p2 <- predict.moult(mmf, newdata = data.frame(day, ssex), predict.type = "start")) (p3 <- predict.moult(mmf, newdata = data.frame(day, ssex), predict.type = "both")) ```

predict.moult function