ConPwrNonMixExp {CP} | R Documentation |

Calculates the conditional power within the non-mixture model with exponential survival.

ConPwrNonMixExp(data, cont.time, new.pat = c(0, 0), theta.0 = 1, alpha = 0.05, disp.data = FALSE, plot.km = FALSE)

`data` |
Data frame which consists of at least three columns with the group (two different expressions) in the first, status (1 = event, 0 = censored) in the second and event time in the third column. |

`cont.time` |
Period of time of continuing the trial. |

`new.pat` |
2-dimensional vector which consists of numbers of new patients who will be recruited each time unit (first component = group 1, second component = group 2) with default at (0, 0). |

`theta.0` |
Originally postulated clinically relevant difference (hazard ratio = hazard of group 2 / hazard of group 1) with default at 1. |

`alpha` |
Significance level for conditional power calculations with default at 0.05. |

`disp.data` |
Logical value indicating if all calculated data should be displayed with default at FALSE. |

`plot.km` |
Logical value indicating if Kaplan-Meier curves and estimated survival curves according to the non-mixture model with exponential survival should be plotted with default at FALSE. |

This function calculates the conditional power within the non-mixture model with exponential survival, i. e.

*S(t) = c^(1 - e^(- λ t))*

for all *t ≥ 0*, *λ > 0* and *0 < c < 1*,
and plots the conditional power curve.

Optionally, further data will be displayed. This includes data from interim analysis, log-likelihoods, AICs, calculated estimators and further patient times.

Moreover, it is possible to plot the Kaplan-Meier curves and the estimated survival curves according to the non-mixture model with exponential survival.

See Details.

Returns a list which consists of the following components:

`lambda1.hat` |
estimated rate parameter of group 1 |

`c1.hat` |
estimated survival fraction of group 1 |

`lambda2.hat` |
estimated rate parameter of group 2 |

`c2.hat` |
estimated survival fraction of group 2 |

`theta.hat` |
estimated hazard ratio
= |

`gamma.theta.0` |
conditional power |

There are several mechanisms to ensure that no illegal operations will be done and maximum likelihood calculations will be executed stable. That is why there should not be too less data in the data frame, for example one patient of each group and both being censored.

Andreas Kuehnapfel

Kuehnapfel, A. (2013). Die bedingte Power in der Ueberlebenszeitanalyse.

# data frame 'test' generated by 'GenerateDataFrame' # conditional power calculations # within the non-mixture model with exponential survival ConPwrNonMixExp(data = test, cont.time = 12, new.pat = c(2.5, 2.5), theta.0 = 0.75, alpha = 0.05, disp.data = TRUE, plot.km = TRUE)

[Package *CP* version 1.6 Index]