Statistical Analysis Stack Exchange
Statistical Analysis Stack Exchange

Questions tagged [segmented-regression]

Use for regression where a different model is fitted for parts of the range of a predictor.

In regression the same relationship may not necessarily be the case over the whole range of a predictor variable. For instance if the predictor is time then some event may have occurred after which the slope of the regression changes. Such a model is also referred to as a broken stick model. In more complex cases it is more than just the slope which changes.

54 questions
8
votes
2 answers

Using segmented linear regression as evidence for the limit of human lifespan

Nature published this year the following paper: Evidence for a limit to human lifespan1, in which the authors argue their "results strongly suggest that the maximum lifespan of humans is fixed and subject to natural constraints." One of this paper's…
Firebug
  • 15,262
  • 5
  • 60
  • 127
7
votes
2 answers

Choosing the number of breakpoints for segmented regression (in R)

What methods/functions are available for choosing the number of breakpoints for segmented regression? Toy data/code for…
Tal Galili
  • 19,935
  • 32
  • 133
  • 195
7
votes
1 answer

Estimate number of breakpoints in regression

I've been playing around with the package strucchange (and to some extent segmented) in R. I'm trying to determine whether there are changes in slope in a linear regression and more importantly, how many breakpoints. A toy dataset: x <- c(0, 5, 10,…
user2414840
  • 263
  • 1
  • 3
  • 4
6
votes
2 answers

R - how to get standard error for a breakpoint/parameter in a piecewise regression

I'm new to asking questions here, so I hope this is a reasonable place for this. I am trying to fit a piece-wise regression. I expect my response (y) to increase from x0 to somewhere along x, then to plateau after a breakpoint; thereafter, slope =0.…
Sarah
  • 81
  • 6
5
votes
0 answers

Segmented regression in R: interpreting the other coefficients

I am running segmented regression using the R package 'segmented'. The original binomial logistic regression has two coefficients, approach_km (continuous), and sea (dichotomous) that explain the probability of the dependent variable occurring. I am…
marcellt
  • 401
  • 5
  • 14
5
votes
1 answer

Statistical significance of breakpoints from `segmented` in R

I'm looking for breakpoints in species abundance as a function of spatial distance, using the segmented package for R. 'segmented' appears to return a breakpoint no matter what; I don't understand whether it returns an estimate of the signifcance of…
Drew Steen
  • 674
  • 6
  • 18
5
votes
1 answer

interrupted time series in R

I have Malaria incidence data from 2003 to 2013. There was intervention implemented in 2008. How can I do segmented regression analysis of interrupted time series in R to test whether the pre intervention slopes are different from post-intervention…
Ambarish Dutta
  • 51
  • 1
  • 2
4
votes
1 answer

What are the pros and cons of segmented regression and regression with ARIMA errors for interrupted time-series analysis?

What are the pros and cons of segmented regression and regression with ARIMA errors for interrupted time-series analysis? In what situation would I choose one method over the other? Are there any papers comparing the two techniques?
fmark
  • 4,666
  • 5
  • 35
  • 51
3
votes
1 answer

Adding additional terms to a piecewise regression

I am exploring the probability of flight in a seabird (1=flight, 0=no flight) using binomial logistic regression. My predictors are distance to a disturbance (continuous), hour of the day (continuous), site (factor), season (factor), sea state…
marcellt
  • 401
  • 5
  • 14
3
votes
1 answer

AIC/BIC for a segmented regression model?

Can AIC/BIC be calculated reliably for a segmented regression model? If so, can the results be compared against AIC/BIC calculations for non-segmented regression models on the same data?
dbspon
  • 119
  • 7
3
votes
3 answers

segmentation of univariate irregular time series

this is my first post. I have an irregular time series that exhibits large shifts in both mean and in the direction of the trend. It looks something like this (though this is far cleaner than reality): set.seed(1) x =…
HEITZ
  • 1,682
  • 7
  • 15
3
votes
1 answer

Help constructing a simple regression model with a breakpoint

This is related to my questions here and here. I am still struggling with my model, so I am taking it back to basics. My assertion is simple, I believe that watershed runoff will have a different relationship with precipitation depending on if the…
Iceberg Slim
  • 261
  • 2
  • 11
2
votes
0 answers

What is the correct specification for a segmented regression of interrupted time series data with a non-equivalent comparison group?

I know this is the general specification for a single group: $Y_i$ = $\beta_0$ + $\beta_1$*time + $\beta_2$*post + $\beta_3$*timepost where time is a continuous variable indicating the time (e.g., month) of every observation in the study; post is an…
Eric Green
  • 629
  • 9
  • 20
2
votes
1 answer

mgcv::gamm extract breakpoints

I want to model the relationship between individuals' weekly mean speed and the timing in the year (represented by the week number), to then extract the dates where we observe a change in the animals' behaviour. I have several ID, accross multiple…
airC
  • 21
  • 2
2
votes
0 answers

Interrupted time series using GLMM / GAMM: which time variables should be smoothed?

I have conducted a controlled interrupted time series analysis within a GLMM framework. I have run the analysis in both the control and treatment groups separately to quantify the basic effects: # Treatment group its.trt <- glmmTMB(out ~ ar1(time +…
C_H
  • 95
  • 8
1
2 3 4