absoluteRisk.CompRisk {casebase} | R Documentation |

Using the output of the function `fitSmoothHazard`

, we can compute
absolute risks by integrating the fitted hazard function over a time period
and then converting this to an estimated survival for each individual.

Plot method for objects returned by the `absoluteRisk`

function. Current plot types are cumulative incidence and survival
functions.

absoluteRisk.CompRisk( object, time, newdata, method = c("numerical", "montecarlo"), nsamp = 100, onlyMain = TRUE, type = c("CI", "survival"), addZero = TRUE ) absoluteRisk( object, time, newdata, method = c("numerical", "montecarlo"), nsamp = 100, s = c("lambda.1se", "lambda.min"), n.trees, onlyMain = TRUE, type = c("CI", "survival"), addZero = TRUE, ntimes = 100, ... ) ## S3 method for class 'absRiskCB' print(x, ...) ## S3 method for class 'absRiskCB' plot( x, ..., xlab = "time", ylab = ifelse(attr(x, "type") == "CI", "cumulative incidence", "survival probability"), type = "l", gg = TRUE, id.names, legend.title )

`object` |
Output of function |

`time` |
A vector of time points at which we should compute the absolute risks. |

`newdata` |
Optionally, a data frame in which to look for variables with
which to predict. If omitted, the mean absolute risk is returned.
Alternatively, if |

`method` |
Method used for integration. Defaults to |

`nsamp` |
Maximal number of subdivisions (if |

`onlyMain` |
Logical. For competing risks, should we return absolute risks
only for the main event of interest? Defaults to |

`type` |
Line type. Only used if |

`addZero` |
Logical. Should we add time = 0 at the beginning of the
output? Defaults to |

`s` |
Value of the penalty parameter lambda at which predictions are
required (for class |

`n.trees` |
Number of trees used in the prediction (for class |

`ntimes` |
Number of time points (only used if |

`...` |
further arguments passed to |

`x` |
Fitted object of class |

`xlab` |
xaxis label, Default: 'time' |

`ylab` |
yaxis label. By default, this will use the |

`gg` |
Logical for whether the |

`id.names` |
Optional character vector used as legend key when |

`legend.title` |
Optional character vector of the legend title. Only used
if |

If `newdata = "typical"`

, we create a typical covariate profile for the
absolute risk computation. This means that we take the median for numerical
and date variables, and we take the most common level for factor variables.

In general, the output will include a column corresponding to the provided
time points. Some modifications of the `time`

vector are done:
`time=0`

is added, the time points are ordered, and duplicates are
removed. All these modifications simplify the computations and give an output
that can easily be used to plot risk curves.

If there is no competing risk, the output is a matrix where each column corresponds to the several covariate profiles, and where each row corresponds to a time point. If there are competing risks, the output will be a 3-dimensional array, with the third dimension corresponding to the different events.

The numerical method should be good enough in most situation, but Monte Carlo integration can give more accurate results when the estimated hazard function is not smooth (e.g. when modeling with time-varying covariates).

If `time`

was provided, returns the estimated absolute risk for
the user-supplied covariate profiles. This will be stored in a matrix or a
higher dimensional array, depending on the input (see details). If both
`time`

and `newdata`

are missing, returns the original data
with a new column containing the risk estimate at failure times.

A plot of the cumulative incidence or survival curve

`matplot`

,
`absoluteRisk`

,
`as.data.table`

, `setattr`

,
`melt.data.table`

# Simulate censored survival data for two outcome types library(data.table) set.seed(12345) nobs <- 1000 tlim <- 20 # simulation parameters b1 <- 200 b2 <- 50 # event type 0-censored, 1-event of interest, 2-competing event # t observed time/endpoint # z is a binary covariate DT <- data.table(z = rbinom(nobs, 1, 0.5)) DT[,`:=` ("t_event" = rweibull(nobs, 1, b1), "t_comp" = rweibull(nobs, 1, b2))] DT[,`:=`("event" = 1 * (t_event < t_comp) + 2 * (t_event >= t_comp), "time" = pmin(t_event, t_comp))] DT[time >= tlim, `:=`("event" = 0, "time" = tlim)] out_linear <- fitSmoothHazard(event ~ time + z, DT, ratio = 10) linear_risk <- absoluteRisk(out_linear, time = 10, newdata = data.table("z" = c(0,1))) # Plot CI curves---- library(ggplot2) data("brcancer") mod_cb_tvc <- fitSmoothHazard(cens ~ estrec*log(time) + horTh + age + menostat + tsize + tgrade + pnodes + progrec, data = brcancer, time = "time", ratio = 1) smooth_risk_brcancer <- absoluteRisk(object = mod_cb_tvc, newdata = brcancer[c(1,50),]) class(smooth_risk_brcancer) plot(smooth_risk_brcancer)

[Package *casebase* version 0.9.1 Index]