These examples illustrate which models, engines, and prediction types are available in poissonreg. As a reminder, in parsnip,
the model type differentiates basic modeling approaches, such as random forests, proportional hazards models, etc.,
the mode denotes in what kind of modeling context it will be used (here: regression), and
the computational engine indicates how the model is fit, such as with a specific R package implementation or even methods outside of R like Keras or Stan.
The following examples use the same data set throughout.
poisson_reg() models
With the "glm" engine
We’ll model the number of articles published by graduate students in biochemistry Ph.D. programs.
## ── Attaching packages ───────────────────────────── tidymodels 1.4.1 ── ## ✔ broom 1.0.12 ✔ rsample 1.3.2
## ✔ dials 1.4.3 ✔ tailor 0.1.0
## ✔ dplyr 1.2.1 ✔ tidyr 1.3.2
## ✔ infer 1.1.0 ✔ tune 2.1.0
## ✔ modeldata 1.5.1 ✔ workflows 1.3.0
## ✔ parsnip 1.5.0 ✔ workflowsets 1.1.1
## ✔ purrr 1.2.2 ✔ yardstick 1.4.0
## ✔ recipes 1.3.2 ## ── Conflicts ──────────────────────────────── tidymodels_conflicts() ──
## ✖ purrr::discard() masks scales::discard()
## ✖ dplyr::filter() masks stats::filter()
## ✖ dplyr::lag() masks stats::lag()
## ✖ recipes::step() masks stats::step()
library(poissonreg)
tidymodels_prefer()
data(bioChemists, package = "pscl")
biochemists_train <- bioChemists[-c(1:5), ]
biochemists_test <- bioChemists[1:5, ]We can define the model:
pr_spec <-
poisson_reg() |>
set_engine("glm") |>
set_mode("regression")
pr_spec ## Poisson Regression Model Specification (regression)
##
## Computational engine: glmNow we create the model fit object:
## parsnip model object
##
##
## Call: stats::glm(formula = art ~ ., family = stats::poisson, data = data)
##
## Coefficients:
## (Intercept) femWomen marMarried kid5 phd
## 0.32187 -0.22207 0.14916 -0.18580 0.01003
## ment
## 0.02559
##
## Degrees of Freedom: 909 Total (i.e. Null); 904 Residual
## Null Deviance: 1800
## Residual Deviance: 1618 AIC: 3298The holdout data can be predicted:
predict(pr_fit, biochemists_test) ## # A tibble: 5 × 1
## .pred
## <dbl>
## 1 1.96
## 2 1.32
## 3 1.34
## 4 1.45
## 5 2.23
With the "glmnet" engine
We’ll model the number of articles published by graduate students in biochemistry Ph.D. programs.
library(tidymodels)
library(poissonreg)
tidymodels_prefer()
data(bioChemists, package = "pscl")
biochemists_train <- bioChemists[-c(1:5), ]
biochemists_test <- bioChemists[1:5, ]We can define the model with specific parameters:
pr_spec <-
poisson_reg(penalty = 0.1) |>
set_engine("glmnet") |>
set_mode("regression")
pr_spec ## Poisson Regression Model Specification (regression)
##
## Main Arguments:
## penalty = 0.1
##
## Computational engine: glmnetNow we create the model fit object:
## parsnip model object
##
##
## Call: glmnet::glmnet(x = maybe_matrix(x), y = y, family = "poisson")
##
## Df %Dev Lambda
## 1 0 0.00 0.59250
## 2 1 1.71 0.53980
## 3 1 3.03 0.49190
## 4 1 4.06 0.44820
## 5 1 4.88 0.40840
## 6 1 5.52 0.37210
## 7 1 6.04 0.33900
## 8 1 6.45 0.30890
## 9 1 6.78 0.28150
## 10 1 7.04 0.25650
## 11 1 7.26 0.23370
## 12 1 7.43 0.21290
## 13 1 7.58 0.19400
## 14 1 7.69 0.17680
## 15 2 7.80 0.16110
## 16 2 7.99 0.14680
## 17 2 8.14 0.13370
## 18 3 8.33 0.12180
## 19 3 8.58 0.11100
## 20 3 8.79 0.10120
## 21 3 8.96 0.09217
## 22 3 9.11 0.08398
## 23 3 9.22 0.07652
## 24 3 9.32 0.06972
## 25 4 9.45 0.06353
## 26 4 9.57 0.05789
## 27 4 9.67 0.05274
## 28 4 9.74 0.04806
## 29 4 9.81 0.04379
## 30 4 9.87 0.03990
## 31 4 9.91 0.03635
## 32 4 9.95 0.03312
## 33 4 9.98 0.03018
## 34 4 10.01 0.02750
## 35 4 10.03 0.02506
## 36 4 10.05 0.02283
## 37 4 10.06 0.02080
## 38 4 10.07 0.01896
## 39 5 10.08 0.01727
## 40 5 10.09 0.01574
## 41 5 10.10 0.01434
## 42 5 10.11 0.01306
## 43 5 10.11 0.01190
## 44 5 10.12 0.01085
## 45 5 10.12 0.00988
## 46 5 10.13 0.00901
## 47 5 10.13 0.00820
## 48 5 10.13 0.00748
## 49 5 10.13 0.00681
## 50 5 10.13 0.00621
## 51 5 10.14 0.00566
## 52 5 10.14 0.00515
## 53 5 10.14 0.00470
## 54 5 10.14 0.00428
## 55 5 10.14 0.00390
## 56 5 10.14 0.00355
## 57 5 10.14 0.00324
## 58 5 10.14 0.00295
## 59 5 10.14 0.00269
## 60 5 10.14 0.00245
## 61 5 10.14 0.00223
## 62 5 10.14 0.00203The holdout data can be predicted:
predict(pr_fit, biochemists_test) ## # A tibble: 5 × 1
## .pred
## <dbl>
## 1 1.67
## 2 1.51
## 3 1.51
## 4 1.49
## 5 2.36
With the "hurdle" engine
We’ll model the number of articles published by graduate students in biochemistry Ph.D. programs.
library(tidymodels)
library(poissonreg)
tidymodels_prefer()
data(bioChemists, package = "pscl")
biochemists_train <- bioChemists[-c(1:5), ]
biochemists_test <- bioChemists[1:5, ]We can define the model:
pr_spec <-
poisson_reg() |>
set_engine("hurdle") |>
set_mode("regression")
pr_spec ## Poisson Regression Model Specification (regression)
##
## Computational engine: hurdleNow we create the model fit object:
## parsnip model object
##
##
## Call:
## pscl::hurdle(formula = art ~ . | ., data = data)
##
## Count model coefficients (truncated poisson with log link):
## (Intercept) femWomen marMarried kid5 phd
## 0.67114 -0.22858 0.09648 -0.14219 -0.01273
## ment
## 0.01875
##
## Zero hurdle model coefficients (binomial with logit link):
## (Intercept) femWomen marMarried kid5 phd
## 0.28643 -0.24805 0.30873 -0.29088 0.01038
## ment
## 0.08317The holdout data can be predicted:
predict(pr_fit, biochemists_test) ## # A tibble: 5 × 1
## .pred
## <dbl>
## 1 2.02
## 2 1.32
## 3 1.31
## 4 1.45
## 5 2.40
With the "stan" engine
We’ll model the number of articles published by graduate students in biochemistry Ph.D. programs.
library(tidymodels)
library(poissonreg)
tidymodels_prefer()
data(bioChemists, package = "pscl")
biochemists_train <- bioChemists[-c(1:5), ]
biochemists_test <- bioChemists[1:5, ]We can define the model:
pr_spec <-
poisson_reg() |>
set_engine("stan") |>
set_mode("regression")
pr_spec ## Poisson Regression Model Specification (regression)
##
## Computational engine: stanNow we create the model fit object:
##
## SAMPLING FOR MODEL 'count' NOW (CHAIN 1).
## Chain 1:
## Chain 1: Gradient evaluation took 5.1e-05 seconds
## Chain 1: 1000 transitions using 10 leapfrog steps per transition would take 0.51 seconds.
## Chain 1: Adjust your expectations accordingly!
## Chain 1:
## Chain 1:
## Chain 1: Iteration: 1 / 2000 [ 0%] (Warmup)
## Chain 1: Iteration: 200 / 2000 [ 10%] (Warmup)
## Chain 1: Iteration: 400 / 2000 [ 20%] (Warmup)
## Chain 1: Iteration: 600 / 2000 [ 30%] (Warmup)
## Chain 1: Iteration: 800 / 2000 [ 40%] (Warmup)
## Chain 1: Iteration: 1000 / 2000 [ 50%] (Warmup)
## Chain 1: Iteration: 1001 / 2000 [ 50%] (Sampling)
## Chain 1: Iteration: 1200 / 2000 [ 60%] (Sampling)
## Chain 1: Iteration: 1400 / 2000 [ 70%] (Sampling)
## Chain 1: Iteration: 1600 / 2000 [ 80%] (Sampling)
## Chain 1: Iteration: 1800 / 2000 [ 90%] (Sampling)
## Chain 1: Iteration: 2000 / 2000 [100%] (Sampling)
## Chain 1:
## Chain 1: Elapsed Time: 0.251 seconds (Warm-up)
## Chain 1: 0.27 seconds (Sampling)
## Chain 1: 0.521 seconds (Total)
## Chain 1:
##
## SAMPLING FOR MODEL 'count' NOW (CHAIN 2).
## Chain 2:
## Chain 2: Gradient evaluation took 3.2e-05 seconds
## Chain 2: 1000 transitions using 10 leapfrog steps per transition would take 0.32 seconds.
## Chain 2: Adjust your expectations accordingly!
## Chain 2:
## Chain 2:
## Chain 2: Iteration: 1 / 2000 [ 0%] (Warmup)
## Chain 2: Iteration: 200 / 2000 [ 10%] (Warmup)
## Chain 2: Iteration: 400 / 2000 [ 20%] (Warmup)
## Chain 2: Iteration: 600 / 2000 [ 30%] (Warmup)
## Chain 2: Iteration: 800 / 2000 [ 40%] (Warmup)
## Chain 2: Iteration: 1000 / 2000 [ 50%] (Warmup)
## Chain 2: Iteration: 1001 / 2000 [ 50%] (Sampling)
## Chain 2: Iteration: 1200 / 2000 [ 60%] (Sampling)
## Chain 2: Iteration: 1400 / 2000 [ 70%] (Sampling)
## Chain 2: Iteration: 1600 / 2000 [ 80%] (Sampling)
## Chain 2: Iteration: 1800 / 2000 [ 90%] (Sampling)
## Chain 2: Iteration: 2000 / 2000 [100%] (Sampling)
## Chain 2:
## Chain 2: Elapsed Time: 0.253 seconds (Warm-up)
## Chain 2: 0.28 seconds (Sampling)
## Chain 2: 0.533 seconds (Total)
## Chain 2:
##
## SAMPLING FOR MODEL 'count' NOW (CHAIN 3).
## Chain 3:
## Chain 3: Gradient evaluation took 3.3e-05 seconds
## Chain 3: 1000 transitions using 10 leapfrog steps per transition would take 0.33 seconds.
## Chain 3: Adjust your expectations accordingly!
## Chain 3:
## Chain 3:
## Chain 3: Iteration: 1 / 2000 [ 0%] (Warmup)
## Chain 3: Iteration: 200 / 2000 [ 10%] (Warmup)
## Chain 3: Iteration: 400 / 2000 [ 20%] (Warmup)
## Chain 3: Iteration: 600 / 2000 [ 30%] (Warmup)
## Chain 3: Iteration: 800 / 2000 [ 40%] (Warmup)
## Chain 3: Iteration: 1000 / 2000 [ 50%] (Warmup)
## Chain 3: Iteration: 1001 / 2000 [ 50%] (Sampling)
## Chain 3: Iteration: 1200 / 2000 [ 60%] (Sampling)
## Chain 3: Iteration: 1400 / 2000 [ 70%] (Sampling)
## Chain 3: Iteration: 1600 / 2000 [ 80%] (Sampling)
## Chain 3: Iteration: 1800 / 2000 [ 90%] (Sampling)
## Chain 3: Iteration: 2000 / 2000 [100%] (Sampling)
## Chain 3:
## Chain 3: Elapsed Time: 0.253 seconds (Warm-up)
## Chain 3: 0.344 seconds (Sampling)
## Chain 3: 0.597 seconds (Total)
## Chain 3:
##
## SAMPLING FOR MODEL 'count' NOW (CHAIN 4).
## Chain 4:
## Chain 4: Gradient evaluation took 3.4e-05 seconds
## Chain 4: 1000 transitions using 10 leapfrog steps per transition would take 0.34 seconds.
## Chain 4: Adjust your expectations accordingly!
## Chain 4:
## Chain 4:
## Chain 4: Iteration: 1 / 2000 [ 0%] (Warmup)
## Chain 4: Iteration: 200 / 2000 [ 10%] (Warmup)
## Chain 4: Iteration: 400 / 2000 [ 20%] (Warmup)
## Chain 4: Iteration: 600 / 2000 [ 30%] (Warmup)
## Chain 4: Iteration: 800 / 2000 [ 40%] (Warmup)
## Chain 4: Iteration: 1000 / 2000 [ 50%] (Warmup)
## Chain 4: Iteration: 1001 / 2000 [ 50%] (Sampling)
## Chain 4: Iteration: 1200 / 2000 [ 60%] (Sampling)
## Chain 4: Iteration: 1400 / 2000 [ 70%] (Sampling)
## Chain 4: Iteration: 1600 / 2000 [ 80%] (Sampling)
## Chain 4: Iteration: 1800 / 2000 [ 90%] (Sampling)
## Chain 4: Iteration: 2000 / 2000 [100%] (Sampling)
## Chain 4:
## Chain 4: Elapsed Time: 0.241 seconds (Warm-up)
## Chain 4: 0.308 seconds (Sampling)
## Chain 4: 0.549 seconds (Total)
## Chain 4:
pr_fit ## parsnip model object
##
## stan_glm
## family: poisson [log]
## formula: art ~ .
## observations: 910
## predictors: 6
## ------
## Median MAD_SD
## (Intercept) 0.3 0.1
## femWomen -0.2 0.1
## marMarried 0.2 0.1
## kid5 -0.2 0.0
## phd 0.0 0.0
## ment 0.0 0.0
##
## ------
## * For help interpreting the printed output see ?print.stanreg
## * For info on the priors used see ?prior_summary.stanregThe holdout data can be predicted:
predict(pr_fit, biochemists_test) ## # A tibble: 5 × 1
## .pred
## <dbl>
## 1 0.674
## 2 0.271
## 3 0.289
## 4 0.371
## 5 0.800
With the "zeroinfl" engine
We’ll model the number of articles published by graduate students in biochemistry Ph.D. programs.
library(tidymodels)
library(poissonreg)
tidymodels_prefer()
data(bioChemists, package = "pscl")
biochemists_train <- bioChemists[-c(1:5), ]
biochemists_test <- bioChemists[1:5, ]We can define the model:
pr_spec <-
poisson_reg() |>
set_engine("zeroinfl") |>
set_mode("regression")
pr_spec ## Poisson Regression Model Specification (regression)
##
## Computational engine: zeroinflNow we create the model fit object:
## parsnip model object
##
##
## Call:
## pscl::zeroinfl(formula = art ~ . | ., data = data)
##
## Count model coefficients (poisson with log link):
## (Intercept) femWomen marMarried kid5 phd
## 0.637600 -0.205020 0.099136 -0.143444 -0.005037
## ment
## 0.018101
##
## Zero-inflation model coefficients (binomial with logit link):
## (Intercept) femWomen marMarried kid5 phd
## -0.65529 0.12466 -0.35861 0.22921 0.03238
## ment
## -0.14778The holdout data can be predicted:
predict(pr_fit, biochemists_test) ## # A tibble: 5 × 1
## .pred
## <dbl>
## 1 2.05
## 2 1.35
## 3 1.32
## 4 1.46
## 5 2.39