Ceteris Paribus Plots

We will explain prediction for

##      m2.price construction.year surface floor no.rooms    district
## 1001     4644              1976     131     3        5 Srodmiescie
wi_rf <- ceteris_paribus(explainer_rf, observation = new_apartment)
wi_rf
##       y_hat new_x             vname   x_quant quant relative_quant
## 0% 4255.354  1920 construction.year 0.6268889  0.00     -0.6268889
## 1% 4300.702  1921 construction.year 0.6268889  0.01     -0.6168889
## 2% 4301.926  1922 construction.year 0.6268889  0.02     -0.6068889
## 3% 4305.352  1923 construction.year 0.6268889  0.03     -0.5968889
## 4% 4305.352  1923 construction.year 0.6268889  0.04     -0.5868889
## 5% 4267.723  1924 construction.year 0.6268889  0.05     -0.5768889
##           label
## 0% randomForest
## 1% randomForest
## 2% randomForest
## 3% randomForest
## 4% randomForest
## 5% randomForest

wi_lm <- ceteris_paribus(explainer_lm, observation = new_apartment)
wi_lm
##           y_hat new_x             vname   x_quant quant relative_quant
## 1001   4832.833  1920 construction.year 0.6268889  0.00     -0.6268889
## 1001.1 4832.604  1921 construction.year 0.6268889  0.01     -0.6168889
## 1001.2 4832.375  1922 construction.year 0.6268889  0.02     -0.6068889
## 1001.3 4832.146  1923 construction.year 0.6268889  0.03     -0.5968889
## 1001.4 4832.146  1923 construction.year 0.6268889  0.04     -0.5868889
## 1001.5 4831.917  1924 construction.year 0.6268889  0.05     -0.5768889
##        label
## 1001      lm
## 1001.1    lm
## 1001.2    lm
## 1001.3    lm
## 1001.4    lm
## 1001.5    lm

Interactive Ceteris Paribus Plots

Interactive Ceteris Paribus Plots - two models