Learning Curve¶

In [17]:

%matplotlib inline
from verticapy.learn.ensemble import RandomForestClassifier
model = RandomForestClassifier(name = "public.RF_titanic")

from verticapy.learn.model_selection import learning_curve

# Efficiency
learning_curve(model,
               input_relation = "public.titanic", 
               X = ["age", "fare", "parch",],
               y = "survived",
               method = "efficiency",
               cv = 3,
               metric = "auc",)

Out[17]:

	n	auc	auc_std	auc_train	auc_train_std	time	time_std
1	122	0.6680075733997303	0.05332017908110594	0.8909149006955003	0.02131965500475238	1.503541072209676	0.37432754554226183
2	397	0.6547157090682778	0.007713486570792588	0.844154459390894	0.0030719106038539327	1.3460146586100261	0.14537931288482728
3	681	0.6845369670886108	0.025313805479825497	0.7927344398212945	0.005728105254458338	1.1761994361877441	0.11630584153901158
4	955	0.6962737437579959	0.0353512031224939	0.7894703543073566	0.01114052118800252	1.062860329945882	0.06530287158725305
5	1234	0.6896610955247162	0.017315622280739237	0.7730183920840378	0.012105548242507798	1.1337089538574219	0.12134504064270113

Rows: 1-5 | Columns: 7

In [18]:

# Scalability
learning_curve(model,
               input_relation = "public.titanic", 
               X = ["age", "fare", "parch",],
               y = "survived",
               method = "scalability",
               cv = 3,
               metric = "auc",)

Out[18]:

	n	auc	auc_std	auc_train	auc_train_std	time	time_std
1	123	0.6814825483859573	0.09712858965790638	0.937303962797384	0.024166714554595854	1.2533116340637207	0.09617902187426004
2	412	0.5890738507854398	0.0024133954753049035	0.8065463750240173	0.014036901839973944	1.3816702365875244	0.25396854032294836
3	683	0.6183559830176371	0.019983410417615116	0.8157142685356973	0.01418662125334137	1.329837163289388	0.13701458312960937
4	971	0.6948259810963078	0.0297371897794578	0.757777896022192	0.013512447824522805	1.2793657779693604	0.10470497909591264
5	1234	0.716133434272017	0.009834874669728913	0.7565986491099747	0.004101615442895625	1.3786905606587727	0.2259709044529702

Rows: 1-5 | Columns: 7

In [19]:

# Performance
learning_curve(model,
               input_relation = "public.titanic", 
               X = ["age", "fare", "parch",],
               y = "survived",
               method = "performance",
               cv = 3,
               metric = "auc",)

Out[19]:

	n	auc	auc_std	auc_train	auc_train_std	time	time_std
1	966	0.7194076143331142	0.018205640397081297	0.7932483691885235	0.004976362571193478	1.1239941120147705	0.0686964846567868
2	674	0.6691529181621174	0.02098854340575278	0.7969280658633137	0.017287338598212996	1.2031710942586262	0.07102441871768375
3	1234	0.6872676450272832	0.028812194076872598	0.7657448410537526	0.006676416147246716	1.2820297082265217	0.3785096907736215
4	422	0.630111446594981	0.060153096299933755	0.8611947560443024	0.00624243924732973	1.5535437266031902	0.4046288796203415
5	121	0.5397620222764927	0.15836977132199492	0.9094177752072491	0.026485218513030862	1.810259183247884	0.893707415201272

Rows: 1-5 | Columns: 7