Applies a modeling function to every combination of a set of formulas and a set of data subsets.
Arguments
- data
A data frame
- formulas
A list of formulas to apply to each subset of the data. If named, these names will be used in the
modelcolumn of the output. Otherwise, the formulas will be converted to strings in themodelcolumn.- cols
Columns to subset the data. Can be any expression supported by <
tidy-select>. IfNULL, the data is not subset into columns. Defaults toNULL.- weights
A list of columns passed to
weightsinfn. If one of the elements isNULLorNA, that model will not be weighted. Defaults toNULL.- clusters
A list of columns passed to
clustersif supported byfn. If one of the elements isNULLorNA, that model will not be clustered. Defaults toNULL.- families
A list of glm model families passed to
familyif supported byfn. Defaults togaussian("identity"), the equivalent oflm(). See family for examples.- fn
The modeling function. Either an unquoted function name or a purrr-style lambda function with two arguments. To use multiple modeling functions, see
cross_fit_glm(). Defaults to lm.- fn_args
A list of additional arguments to
fn.- tidy
A logical or function to use to tidy model output into data.frame columns. If
TRUE, uses the default tidying function:tidy_glance(). IfFALSE,NA, orNULL, the untidied model output will be returned in a list column namedfit. An alternative function can be specified with an unquoted function name or a purrr-style lambda function with one argument (see usage with broom::tidy(conf.int = TRUE) in examples). Defaults to tidy_glance.- tidy_args
A list of additional arguments to the
tidyfunction- errors
If
"stop", the default, the function will stop and return an error if any subset produces an error. If"warn", the function will produce a warning for subsets that produce an error and return results for all subsets that do not.
Value
A tibble with a column for the model formula,
columns for subsets,
columns for the model family and type (if applicable),
columns for the weights and clusters (if applicable),
and columns of tidy model output or a list column of models
(if tidy = FALSE)
See also
cross_fit_glm() to map a model across multiple model types.
cross_fit_robust() to map robust linear models.
xmap() to apply any function to combinations of inputs.
Examples
cross_fit(mtcars, mpg ~ wt, cyl)
#> # A tibble: 6 × 19
#> model cyl term estimate std.error statistic p.value r.squared adj.r.squared
#> <chr> <dbl> <chr> <dbl> <dbl> <dbl> <dbl> <dbl> <dbl>
#> 1 mpg … 4 (Int… 39.6 4.35 9.10 7.77e-6 0.509 0.454
#> 2 mpg … 4 wt -5.65 1.85 -3.05 1.37e-2 0.509 0.454
#> 3 mpg … 6 (Int… 28.4 4.18 6.79 1.05e-3 0.465 0.357
#> 4 mpg … 6 wt -2.78 1.33 -2.08 9.18e-2 0.465 0.357
#> 5 mpg … 8 (Int… 23.9 3.01 7.94 4.05e-6 0.423 0.375
#> 6 mpg … 8 wt -2.19 0.739 -2.97 1.18e-2 0.423 0.375
#> # ℹ 10 more variables: sigma <dbl>, model.statistic <dbl>, model.p.value <dbl>,
#> # df <dbl>, logLik <dbl>, AIC <dbl>, BIC <dbl>, deviance <dbl>,
#> # df.residual <int>, nobs <int>
cross_fit(mtcars, list(mpg ~ wt, mpg ~ hp), cyl)
#> # A tibble: 12 × 19
#> model cyl term estimate std.error statistic p.value r.squared
#> <chr> <dbl> <chr> <dbl> <dbl> <dbl> <dbl> <dbl>
#> 1 mpg ~ wt 4 (Intercept) 39.6 4.35 9.10 0.00000777 0.509
#> 2 mpg ~ wt 4 wt -5.65 1.85 -3.05 0.0137 0.509
#> 3 mpg ~ wt 6 (Intercept) 28.4 4.18 6.79 0.00105 0.465
#> 4 mpg ~ wt 6 wt -2.78 1.33 -2.08 0.0918 0.465
#> 5 mpg ~ wt 8 (Intercept) 23.9 3.01 7.94 0.00000405 0.423
#> 6 mpg ~ wt 8 wt -2.19 0.739 -2.97 0.0118 0.423
#> 7 mpg ~ hp 4 (Intercept) 36.0 5.20 6.92 0.0000693 0.274
#> 8 mpg ~ hp 4 hp -0.113 0.0612 -1.84 0.0984 0.274
#> 9 mpg ~ hp 6 (Intercept) 20.7 3.30 6.26 0.00153 0.0161
#> 10 mpg ~ hp 6 hp -0.00761 0.0266 -0.286 0.786 0.0161
#> 11 mpg ~ hp 8 (Intercept) 18.1 2.99 6.05 0.0000574 0.0804
#> 12 mpg ~ hp 8 hp -0.0142 0.0139 -1.02 0.326 0.0804
#> # ℹ 11 more variables: adj.r.squared <dbl>, sigma <dbl>, model.statistic <dbl>,
#> # model.p.value <dbl>, df <dbl>, logLik <dbl>, AIC <dbl>, BIC <dbl>,
#> # deviance <dbl>, df.residual <int>, nobs <int>
cross_fit(mtcars, list(wt = mpg ~ wt, hp = mpg ~ hp), cyl)
#> # A tibble: 12 × 19
#> model cyl term estimate std.error statistic p.value r.squared
#> <chr> <dbl> <chr> <dbl> <dbl> <dbl> <dbl> <dbl>
#> 1 wt 4 (Intercept) 39.6 4.35 9.10 0.00000777 0.509
#> 2 wt 4 wt -5.65 1.85 -3.05 0.0137 0.509
#> 3 wt 6 (Intercept) 28.4 4.18 6.79 0.00105 0.465
#> 4 wt 6 wt -2.78 1.33 -2.08 0.0918 0.465
#> 5 wt 8 (Intercept) 23.9 3.01 7.94 0.00000405 0.423
#> 6 wt 8 wt -2.19 0.739 -2.97 0.0118 0.423
#> 7 hp 4 (Intercept) 36.0 5.20 6.92 0.0000693 0.274
#> 8 hp 4 hp -0.113 0.0612 -1.84 0.0984 0.274
#> 9 hp 6 (Intercept) 20.7 3.30 6.26 0.00153 0.0161
#> 10 hp 6 hp -0.00761 0.0266 -0.286 0.786 0.0161
#> 11 hp 8 (Intercept) 18.1 2.99 6.05 0.0000574 0.0804
#> 12 hp 8 hp -0.0142 0.0139 -1.02 0.326 0.0804
#> # ℹ 11 more variables: adj.r.squared <dbl>, sigma <dbl>, model.statistic <dbl>,
#> # model.p.value <dbl>, df <dbl>, logLik <dbl>, AIC <dbl>, BIC <dbl>,
#> # deviance <dbl>, df.residual <int>, nobs <int>
cross_fit(mtcars, list(mpg ~ wt, mpg ~ hp), c(cyl, vs))
#> Warning: There were 4 warnings in `dplyr::reframe()`.
#> The first warning was:
#> ℹ In argument: `tidy(...)`.
#> ℹ In row 8.
#> Caused by warning in `summary.lm()`:
#> ! essentially perfect fit: summary may be unreliable
#> ℹ Run `dplyr::last_dplyr_warnings()` to see the 3 remaining warnings.
#> # A tibble: 20 × 20
#> model cyl vs term estimate std.error statistic p.value r.squared
#> <chr> <dbl> <dbl> <chr> <dbl> <dbl> <dbl> <dbl> <dbl>
#> 1 mpg ~ … 4 0 (Int… 26 NaN NaN NaN 0
#> 2 mpg ~ … 4 0 wt NA NA NA NA 0
#> 3 mpg ~ … 4 1 (Int… 39.9 4.61e+ 0 8.66e+ 0 2.47e- 5 0.520
#> 4 mpg ~ … 4 1 wt -5.72 1.95e+ 0 -2.94e+ 0 1.87e- 2 0.520
#> 5 mpg ~ … 6 0 (Int… 22.2 1.61e+ 1 1.38e+ 0 3.99e- 1 0.0103
#> 6 mpg ~ … 6 0 wt -0.594 5.83e+ 0 -1.02e- 1 9.35e- 1 0.0103
#> 7 mpg ~ … 6 1 (Int… 63.6 1.19e+ 1 5.36e+ 0 3.30e- 2 0.876
#> 8 mpg ~ … 6 1 wt -13.1 3.50e+ 0 -3.75e+ 0 6.42e- 2 0.876
#> 9 mpg ~ … 8 0 (Int… 23.9 3.01e+ 0 7.94e+ 0 4.05e- 6 0.423
#> 10 mpg ~ … 8 0 wt -2.19 7.39e- 1 -2.97e+ 0 1.18e- 2 0.423
#> 11 mpg ~ … 4 0 (Int… 26 NaN NaN NaN 0
#> 12 mpg ~ … 4 0 hp NA NA NA NA 0
#> 13 mpg ~ … 4 1 (Int… 36.0 5.52e+ 0 6.52e+ 0 1.85e- 4 0.273
#> 14 mpg ~ … 4 1 hp -0.113 6.55e- 2 -1.73e+ 0 1.21e- 1 0.273
#> 15 mpg ~ … 6 0 (Int… 23.2 1.02e-14 2.28e+15 2.79e-16 1
#> 16 mpg ~ … 6 0 hp -0.0200 7.53e-17 -2.66e+14 2.40e-15 1
#> 17 mpg ~ … 6 1 (Int… 24.2 1.41e+ 1 1.72e+ 0 2.28e- 1 0.0613
#> 18 mpg ~ … 6 1 hp -0.0440 1.22e- 1 -3.61e- 1 7.52e- 1 0.0613
#> 19 mpg ~ … 8 0 (Int… 18.1 2.99e+ 0 6.05e+ 0 5.74e- 5 0.0804
#> 20 mpg ~ … 8 0 hp -0.0142 1.39e- 2 -1.02e+ 0 3.26e- 1 0.0804
#> # ℹ 11 more variables: adj.r.squared <dbl>, sigma <dbl>, model.statistic <dbl>,
#> # model.p.value <dbl>, df <dbl>, logLik <dbl>, AIC <dbl>, BIC <dbl>,
#> # deviance <dbl>, df.residual <int>, nobs <int>
cross_fit(mtcars, list(mpg ~ wt, mpg ~ hp), dplyr::starts_with("c"))
#> # A tibble: 36 × 20
#> model cyl carb term estimate std.error statistic p.value r.squared
#> <chr> <dbl> <dbl> <chr> <dbl> <dbl> <dbl> <dbl> <dbl>
#> 1 mpg ~ wt 4 1 (Inter… 62.0 17.2 3.60 3.67e-2 0.574
#> 2 mpg ~ wt 4 1 wt -16.0 7.95 -2.01 1.38e-1 0.574
#> 3 mpg ~ wt 4 2 (Inter… 36.8 2.83 13.0 2.01e-4 0.802
#> 4 mpg ~ wt 4 2 wt -4.56 1.13 -4.02 1.59e-2 0.802
#> 5 mpg ~ wt 6 1 (Inter… 64.7 NaN NaN NaN 1
#> 6 mpg ~ wt 6 1 wt -13.5 NaN NaN NaN 1
#> 7 mpg ~ wt 6 4 (Inter… 30.2 3.61 8.37 1.40e-2 0.810
#> 8 mpg ~ wt 6 4 wt -3.38 1.16 -2.92 1.00e-1 0.810
#> 9 mpg ~ wt 6 6 (Inter… 19.7 NaN NaN NaN 0
#> 10 mpg ~ wt 6 6 wt NA NA NA NA 0
#> # ℹ 26 more rows
#> # ℹ 11 more variables: adj.r.squared <dbl>, sigma <dbl>, model.statistic <dbl>,
#> # model.p.value <dbl>, df <dbl>, logLik <dbl>, AIC <dbl>, BIC <dbl>,
#> # deviance <dbl>, df.residual <int>, nobs <int>
cross_fit(mtcars, list(hp = mpg ~ hp), cyl, weights = wt)
#> # A tibble: 6 × 20
#> model weights cyl term estimate std.error statistic p.value r.squared
#> <chr> <chr> <dbl> <chr> <dbl> <dbl> <dbl> <dbl> <dbl>
#> 1 hp wt 4 (Intercept) 36.2 5.14 7.03 6.10e-5 0.316
#> 2 hp wt 4 hp -0.123 0.0601 -2.04 7.17e-2 0.316
#> 3 hp wt 6 (Intercept) 20.4 3.49 5.85 2.07e-3 0.0107
#> 4 hp wt 6 hp -0.00657 0.0283 -0.232 8.26e-1 0.0107
#> 5 hp wt 8 (Intercept) 18.0 3.36 5.36 1.72e-4 0.0724
#> 6 hp wt 8 hp -0.0151 0.0156 -0.968 3.52e-1 0.0724
#> # ℹ 11 more variables: adj.r.squared <dbl>, sigma <dbl>, model.statistic <dbl>,
#> # model.p.value <dbl>, df <dbl>, logLik <dbl>, AIC <dbl>, BIC <dbl>,
#> # deviance <dbl>, df.residual <int>, nobs <int>
cross_fit(mtcars, list(hp = mpg ~ hp), cyl, weights = c(wt, NA))
#> # A tibble: 12 × 20
#> model weights cyl term estimate std.error statistic p.value r.squared
#> <chr> <chr> <dbl> <chr> <dbl> <dbl> <dbl> <dbl> <dbl>
#> 1 hp NA 4 (Intercep… 36.0 5.20 6.92 6.93e-5 0.274
#> 2 hp NA 4 hp -0.113 0.0612 -1.84 9.84e-2 0.274
#> 3 hp NA 6 (Intercep… 20.7 3.30 6.26 1.53e-3 0.0161
#> 4 hp NA 6 hp -0.00761 0.0266 -0.286 7.86e-1 0.0161
#> 5 hp NA 8 (Intercep… 18.1 2.99 6.05 5.74e-5 0.0804
#> 6 hp NA 8 hp -0.0142 0.0139 -1.02 3.26e-1 0.0804
#> 7 hp wt 4 (Intercep… 36.2 5.14 7.03 6.10e-5 0.316
#> 8 hp wt 4 hp -0.123 0.0601 -2.04 7.17e-2 0.316
#> 9 hp wt 6 (Intercep… 20.4 3.49 5.85 2.07e-3 0.0107
#> 10 hp wt 6 hp -0.00657 0.0283 -0.232 8.26e-1 0.0107
#> 11 hp wt 8 (Intercep… 18.0 3.36 5.36 1.72e-4 0.0724
#> 12 hp wt 8 hp -0.0151 0.0156 -0.968 3.52e-1 0.0724
#> # ℹ 11 more variables: adj.r.squared <dbl>, sigma <dbl>, model.statistic <dbl>,
#> # model.p.value <dbl>, df <dbl>, logLik <dbl>, AIC <dbl>, BIC <dbl>,
#> # deviance <dbl>, df.residual <int>, nobs <int>
cross_fit(
mtcars, list(vs ~ cyl, vs ~ hp), am,
fn = glm, fn_args = list(family = binomial(link = logit))
)
#> Warning: There were 4 warnings in `dplyr::reframe()`.
#> The first warning was:
#> ℹ In argument: `tidy(...)`.
#> ℹ In row 1.
#> Caused by warning:
#> ! glm.fit: fitted probabilities numerically 0 or 1 occurred
#> ℹ Run `dplyr::last_dplyr_warnings()` to see the 3 remaining warnings.
#> # A tibble: 8 × 15
#> model am term estimate std.error statistic p.value null.deviance df.null
#> <chr> <dbl> <chr> <dbl> <dbl> <dbl> <dbl> <dbl> <int>
#> 1 vs ~ c… 0 (Int… 172. 2.52e+5 0.000682 0.999 25.0 18
#> 2 vs ~ c… 0 cyl -24.6 3.57e+4 -0.000689 0.999 25.0 18
#> 3 vs ~ c… 1 (Int… 44.3 1.04e+4 0.00424 0.997 17.9 12
#> 4 vs ~ c… 1 cyl -10.6 2.61e+3 -0.00406 0.997 17.9 12
#> 5 vs ~ hp 0 (Int… 231. 2.76e+5 0.000835 0.999 25.0 18
#> 6 vs ~ hp 0 hp -1.69 2.00e+3 -0.000844 0.999 25.0 18
#> 7 vs ~ hp 1 (Int… 7.07 4.76e+0 1.49 0.137 17.9 12
#> 8 vs ~ hp 1 hp -0.0663 4.67e-2 -1.42 0.156 17.9 12
#> # ℹ 6 more variables: logLik <dbl>, AIC <dbl>, BIC <dbl>, deviance <dbl>,
#> # df.residual <int>, nobs <int>
cross_fit(
mtcars, list(vs ~ cyl, vs ~ hp), am,
fn = ~ glm(.x, .y, family = binomial(link = logit))
)
#> Warning: There were 4 warnings in `dplyr::reframe()`.
#> The first warning was:
#> ℹ In argument: `tidy(...)`.
#> ℹ In row 1.
#> Caused by warning:
#> ! glm.fit: fitted probabilities numerically 0 or 1 occurred
#> ℹ Run `dplyr::last_dplyr_warnings()` to see the 3 remaining warnings.
#> # A tibble: 8 × 15
#> model am term estimate std.error statistic p.value null.deviance df.null
#> <chr> <dbl> <chr> <dbl> <dbl> <dbl> <dbl> <dbl> <int>
#> 1 vs ~ c… 0 (Int… 172. 2.52e+5 0.000682 0.999 25.0 18
#> 2 vs ~ c… 0 cyl -24.6 3.57e+4 -0.000689 0.999 25.0 18
#> 3 vs ~ c… 1 (Int… 44.3 1.04e+4 0.00424 0.997 17.9 12
#> 4 vs ~ c… 1 cyl -10.6 2.61e+3 -0.00406 0.997 17.9 12
#> 5 vs ~ hp 0 (Int… 231. 2.76e+5 0.000835 0.999 25.0 18
#> 6 vs ~ hp 0 hp -1.69 2.00e+3 -0.000844 0.999 25.0 18
#> 7 vs ~ hp 1 (Int… 7.07 4.76e+0 1.49 0.137 17.9 12
#> 8 vs ~ hp 1 hp -0.0663 4.67e-2 -1.42 0.156 17.9 12
#> # ℹ 6 more variables: logLik <dbl>, AIC <dbl>, BIC <dbl>, deviance <dbl>,
#> # df.residual <int>, nobs <int>
cross_fit(mtcars, list(mpg ~ wt, mpg ~ hp), cyl, tidy = FALSE)
#> # A tibble: 6 × 3
#> model cyl fit
#> <chr> <dbl> <list>
#> 1 mpg ~ wt 4 <lm>
#> 2 mpg ~ wt 6 <lm>
#> 3 mpg ~ wt 8 <lm>
#> 4 mpg ~ hp 4 <lm>
#> 5 mpg ~ hp 6 <lm>
#> 6 mpg ~ hp 8 <lm>
cross_fit(mtcars, list(mpg ~ wt, mpg ~ hp), cyl, tidy_args = c(conf.int = TRUE))
#> # A tibble: 12 × 21
#> model cyl term estimate std.error statistic p.value conf.low conf.high
#> <chr> <dbl> <chr> <dbl> <dbl> <dbl> <dbl> <dbl> <dbl>
#> 1 mpg ~ wt 4 (Inte… 39.6 4.35 9.10 7.77e-6 29.7 49.4
#> 2 mpg ~ wt 4 wt -5.65 1.85 -3.05 1.37e-2 -9.83 -1.46
#> 3 mpg ~ wt 6 (Inte… 28.4 4.18 6.79 1.05e-3 17.7 39.2
#> 4 mpg ~ wt 6 wt -2.78 1.33 -2.08 9.18e-2 -6.21 0.651
#> 5 mpg ~ wt 8 (Inte… 23.9 3.01 7.94 4.05e-6 17.3 30.4
#> 6 mpg ~ wt 8 wt -2.19 0.739 -2.97 1.18e-2 -3.80 -0.582
#> 7 mpg ~ hp 4 (Inte… 36.0 5.20 6.92 6.93e-5 24.2 47.7
#> 8 mpg ~ hp 4 hp -0.113 0.0612 -1.84 9.84e-2 -0.251 0.0256
#> 9 mpg ~ hp 6 (Inte… 20.7 3.30 6.26 1.53e-3 12.2 29.2
#> 10 mpg ~ hp 6 hp -0.00761 0.0266 -0.286 7.86e-1 -0.0759 0.0607
#> 11 mpg ~ hp 8 (Inte… 18.1 2.99 6.05 5.74e-5 11.6 24.6
#> 12 mpg ~ hp 8 hp -0.0142 0.0139 -1.02 3.26e-1 -0.0445 0.0160
#> # ℹ 12 more variables: r.squared <dbl>, adj.r.squared <dbl>, sigma <dbl>,
#> # model.statistic <dbl>, model.p.value <dbl>, df <dbl>, logLik <dbl>,
#> # AIC <dbl>, BIC <dbl>, deviance <dbl>, df.residual <int>, nobs <int>
cross_fit(mtcars, list(mpg ~ wt, mpg ~ hp), cyl, tidy = broom::tidy)
#> # A tibble: 12 × 7
#> model cyl term estimate std.error statistic p.value
#> <chr> <dbl> <chr> <dbl> <dbl> <dbl> <dbl>
#> 1 mpg ~ wt 4 (Intercept) 39.6 4.35 9.10 0.00000777
#> 2 mpg ~ wt 4 wt -5.65 1.85 -3.05 0.0137
#> 3 mpg ~ wt 6 (Intercept) 28.4 4.18 6.79 0.00105
#> 4 mpg ~ wt 6 wt -2.78 1.33 -2.08 0.0918
#> 5 mpg ~ wt 8 (Intercept) 23.9 3.01 7.94 0.00000405
#> 6 mpg ~ wt 8 wt -2.19 0.739 -2.97 0.0118
#> 7 mpg ~ hp 4 (Intercept) 36.0 5.20 6.92 0.0000693
#> 8 mpg ~ hp 4 hp -0.113 0.0612 -1.84 0.0984
#> 9 mpg ~ hp 6 (Intercept) 20.7 3.30 6.26 0.00153
#> 10 mpg ~ hp 6 hp -0.00761 0.0266 -0.286 0.786
#> 11 mpg ~ hp 8 (Intercept) 18.1 2.99 6.05 0.0000574
#> 12 mpg ~ hp 8 hp -0.0142 0.0139 -1.02 0.326
cross_fit(
mtcars, list(mpg ~ wt, mpg ~ hp), cyl,
tidy = ~ broom::tidy(., conf.int = TRUE)
)
#> # A tibble: 12 × 9
#> model cyl term estimate std.error statistic p.value conf.low conf.high
#> <chr> <dbl> <chr> <dbl> <dbl> <dbl> <dbl> <dbl> <dbl>
#> 1 mpg ~ wt 4 (Inte… 39.6 4.35 9.10 7.77e-6 29.7 49.4
#> 2 mpg ~ wt 4 wt -5.65 1.85 -3.05 1.37e-2 -9.83 -1.46
#> 3 mpg ~ wt 6 (Inte… 28.4 4.18 6.79 1.05e-3 17.7 39.2
#> 4 mpg ~ wt 6 wt -2.78 1.33 -2.08 9.18e-2 -6.21 0.651
#> 5 mpg ~ wt 8 (Inte… 23.9 3.01 7.94 4.05e-6 17.3 30.4
#> 6 mpg ~ wt 8 wt -2.19 0.739 -2.97 1.18e-2 -3.80 -0.582
#> 7 mpg ~ hp 4 (Inte… 36.0 5.20 6.92 6.93e-5 24.2 47.7
#> 8 mpg ~ hp 4 hp -0.113 0.0612 -1.84 9.84e-2 -0.251 0.0256
#> 9 mpg ~ hp 6 (Inte… 20.7 3.30 6.26 1.53e-3 12.2 29.2
#> 10 mpg ~ hp 6 hp -0.00761 0.0266 -0.286 7.86e-1 -0.0759 0.0607
#> 11 mpg ~ hp 8 (Inte… 18.1 2.99 6.05 5.74e-5 11.6 24.6
#> 12 mpg ~ hp 8 hp -0.0142 0.0139 -1.02 3.26e-1 -0.0445 0.0160