Returns tidy data frames of per-stress-level parameter estimates and (when available) global life-stress relationship coefficients from a fitted `alt` object. Useful for reporting ALT results in reproducible research workflows.
Value
A named list with two elements:
- `parallel`
A `data.frame` with one row per stress level, containing columns `stress`, `P1`, `P2`, `wt`, and `n_failures`. For Weibull models, `P1` is the scale parameter \(\eta\) and `P2` is the shape parameter \(\beta\). For lognormal models, `P1` is \(\mu_{log}\) and `P2` is \(\sigma_{log}\).
- `relationship`
A one-row `data.frame` with columns `model`, `coef1`, and `coef2` (the global life-stress relationship coefficients), or `NULL` with a message if `alt.fit()` has not been called.
Details
Two life-stress models are supported. For the Arrhenius model,
\(\eta = \exp(C_1 + C_2 \cdot S)\) where \(S\) is stress (typically
reciprocal temperature in 1/K). For the Power Law model,
\(\eta = \exp(C_1) / S^{|C_2|}\). Both coef1 and coef2
are on the log scale as returned by alt.fit().
References
Nelson, W. B. (1990). Accelerated Testing: Statistical Models, Test Plans, and Data Analysis. Wiley.
See also
[plotly_alt()] for the ALT probability plot, [plotly_rel()] for the life-stress relationship plot.
Examples
library(WeibullR.ALT)
d1 <- alt.data(c(248, 456, 528, 731, 813, 537), stress = 300)
d2 <- alt.data(c(164, 176, 289), stress = 350)
d3 <- alt.data(c(88, 112, 152), stress = 400)
obj <- alt.fit(
alt.parallel(
alt.make(list(d1, d2, d3), dist = "weibull", alt.model = "arrhenius",
view_dist_fits = FALSE),
view_parallel_fits = FALSE
)
)
result <- tidy_alt(obj)
result$parallel
#> stress P1 P2 wt n_failures
#> 1 300 615.9660 3.935945 6 6
#> 2 350 231.8588 3.935945 3 3
#> 3 400 128.0115 3.935945 3 3
result$relationship
#> model coef1 coef2
#> 1 arrhenius 11.20605 -0.01605895