prc_auc¶

In [ ]:

prc_auc(y_true: str, 
        y_score: str, 
        input_relation: Union[str, vDataFrame],
        pos_label: Union[int, float, str] = 1,
        nbins: int = 10000,)

Computes the PRC AUC (Area Under Curve).

Parameters¶

Name	Type	Optional	Description
y_true	str	❌	Response column.
y_score	str	❌	Prediction Probability.
input_relation	str / vDataFrame	❌	The relation used for scoring. This relation can be a view, table, or customized relation. For example, you could write: "(SELECT ... FROM ...) x" as long as an alias is given at the end of the relation.
pos_label	int / float / str	✓	To compute the PRC AUC, one of the response column class has to be the positive one. The parameter 'pos_label' represents this class.
nbins	int	✓	An integer value that determines the number of decision boundaries. Decision boundaries are set at equally spaced intervals between 0 and 1, inclusive. Greater values for nbins give more precise estimations of the AUC, but can potentially decrease performance. The maximum value is 999,999. If negative, the maximum value is used.

Returns¶

float : score

Example¶

In [143]:

from verticapy import vDataFrame
vDataFrame("example_classification")

	123 y_score Float	123 y_true Int	123 y_pred Int
1	0.261992636494471	0	0
2	0.271766949212011	0	0
3	0.281715565124816	0	0
4	0.287704603820825	0	0
5	0.293503745589547	0	0
6	0.294754738264003	0	0
7	0.299562949987817	1	0
8	0.302916327585091	0	0
9	0.304072718408127	0	0
10	0.305575345919111	1	0
11	0.310835669040619	0	0
12	0.310835669040619	1	0
13	0.314677498532068	1	0
14	0.314772537794701	0	0
15	0.317504850884067	0	0
16	0.321047513138065	0	0
17	0.324898318971236	0	0
18	0.327313634871091	0	0
19	0.328290953867949	0	0
20	0.329530097026026	0	0
21	0.330667481508048	0	0
22	0.331694822691007	0	0
23	0.334891289625077	0	0
24	0.339224975430521	0	0
25	0.340649411718069	0	0
26	0.340914231767138	1	0
27	0.34178321869448	0	0
28	0.342456823846109	0	0
29	0.342607538574227	0	0
30	0.343380249983169	1	0
31	0.343380249983169	1	0
32	0.345611509579921	0	0
33	0.345703353743199	0	0
34	0.346341459924855	1	0
35	0.346566377721986	0	0
36	0.34720574465618	0	0
37	0.347320286991246	0	0
38	0.347361879717578	0	0
39	0.347895646927998	0	0
40	0.350041368591872	0	0
41	0.350713668608726	0	0
42	0.351097328234269	1	0
43	0.352016830512644	0	0
44	0.352054220262656	0	0
45	0.352146855196749	0	0
46	0.352239501254866	0	0
47	0.353175137762404	0	0
48	0.3539395916197	1	0
49	0.353996333596281	1	0
50	0.354381624264835	0	0
51	0.354381624264835	0	0
52	0.355342565735203	0	0
53	0.357132185344874	0	0
54	0.357455496335166	0	0
55	0.357679708354009	0	0
56	0.358738033924215	0	0
57	0.359210075590932	0	0
58	0.359566016764265	0	0
59	0.359641574716634	0	0
60	0.36060286105325	0	0
61	0.36060286105325	0	0
62	0.36101075208364	0	0
63	0.362255481975375	0	0
64	0.363178572038573	0	0
65	0.363600418603227	0	0
66	0.364693844779539	0	0
67	0.364907992756202	0	0
68	0.365288848686146	0	0
69	0.367100035711998	0	0
70	0.367262588765524	0	0
71	0.367427213035335	0	0
72	0.368048294763488	0	0
73	0.369225730045058	0	0
74	0.369797716658633	0	0
75	0.370553858267955	0	0
76	0.370553858267955	1	0
77	0.370753825925793	0	0
78	0.37140040117994	1	0
79	0.37207963832586	0	0
80	0.37272228535973	1	0
81	0.37279074038996	0	0
82	0.37337254481298	0	0
83	0.373809604665009	0	0
84	0.373973373195992	1	0
85	0.373973437248784	0	0
86	0.374595381587019	1	0
87	0.376588338299031	0	0
88	0.376653483221958	0	0
89	0.377288922205616	0	0
90	0.378217229733463	0	0
91	0.378825744070643	0	0
92	0.378825744070643	0	0
93	0.378825744070643	1	0
94	0.378825744070643	1	0
95	0.37911132124581	0	0
96	0.380187437748189	0	0
97	0.38033983811501	0	0
98	0.380568066078983	0	0
99	0.380732995250046	0	0
100	0.380876617505906	0	0

Out[143]:

Rows: 1-100 of 1234 | Columns: 3

In [144]:

from verticapy.learn.metrics import prc_auc
prc_auc("y_true", "y_score", "example_classification")

Out[144]:

0.6003540469187277