Modelling flowering of plants using time-to-event methods |
| |
Authors: | Christian Ritz Christian Pipper Flemming Yndgaard Kenneth Fredlund Gerhard Steinrücken |
| |
Institution: | 1. Statistics Group, Department of Basic Sciences and Environment, Faculty of Life Sciences, University of Copenhagen, Thorvaldsensvej 40, DK-1871 Frederiksberg C, Denmark;2. Syngenta Seeds, Box 302, Sockerbruksv. 24, SE-26123 Landskrona, Sweden;1. Department of Mathematics and Statistics, IIT Kanpur, Kanpur 208016, India;2. Dipartimento di Matematica “Giuseppe Peano”, via Carlo Alberto 10, Università di Torino, 10123 Torino, Italy |
| |
Abstract: | Climate and environmental factors are known to influence the propensity of plants to flower. In order to capture the time dynamics in the underlying biological processes, this paper proposes to use a grouped-data survival analysis methodology. We formulate a semi-parametric model for the hazard rate of flowering, which allows all available climatic and environmental data to be included. Estimation can be carried using a generalized linear model approach. In addition, the model can be validated by means of recently developed residual-based model-checking procedures similar to what is available for ordinary regression models. The main finding is that the model proposed captures the major trends and is a first step in the direction of a model-based quantification of some climatic and environmental factors that influence the flowering of plants. |
| |
Keywords: | |
本文献已被 ScienceDirect 等数据库收录! |
|