Result Detail – Random Forest Help

Back to Results Dashboard view Export CSV Export Excel Export PDF Download Full Report

Download Model (.pkl) Notebook

Result Information

Result ID: 149
Type: Random Forest
Status: completed
Dataset ID: 703
Created: 2026-09-21 02:16:12.854253
Saved Model: Available — Download .pkl

Possible data leakage detected:

Identifier-like column(s) used as features: PLOT. These are bookkeeping fields; high importance for them usually reflects row ordering, not biology. Consider excluding them.
Spatial confounding risk: coordinate column(s) LONG rank among the most important features. The model may be learning field position rather than biological signal; interpret with care or exclude coordinates.
Mean cross-validated R² is negative (-0.128) — the model performs worse than simply predicting the mean target value. Results are unreliable; check for leakage, too few samples, or an unsuitable target.

Model Parameters

Model Type: regression
Estimators: 60
Max Depth: None

Pre-Modeling Statistics (Pearson, Spearman, ANOVA) Help

Feature	Pearson r	Pearson p	Pearson FDR	Spearman ρ	Spearman FDR	ANOVA F	ANOVA FDR	Significant?
PLOT	-0.2171	0.0000	0.0001	-0.1240	0.0460	19.3401	0.0001	✓ Yes
ENV_SOILNH4	0.0039	0.9386	0.9386	0.0896	0.1380	0.0059	0.9386	✗ No
ENV_SOILNO3	-0.0257	0.6116	0.6728	-0.0559	0.3171	0.2583	0.6728	✗ No
ENV_SOILP	-0.0515	0.3082	0.3912	-0.0874	0.1380	1.0409	0.3912	✗ No
ENV_SOILOC	0.1615	0.0013	0.0032	0.0991	0.1168	10.4775	0.0032	✓ Yes
ENV_SOILTN	0.1422	0.0047	0.0087	0.0735	0.2008	8.0717	0.0087	✓ Yes
ENV_SOILTC	0.1457	0.0038	0.0078	0.0825	0.1610	8.4802	0.0078	✓ Yes
ENV_PH	0.0094	0.8520	0.8787	-0.0366	0.4841	0.0348	0.8787	✗ No
ENV_SOILCLAY	0.2271	0.0000	0.0000	0.1408	0.0244	21.2587	0.0000	✓ Yes
ENV_SOILSILT	0.2150	0.0000	0.0001	0.0509	0.3451	18.9547	0.0001	✓ Yes
ENV_SOILSAND	-0.2302	0.0000	0.0000	-0.0880	0.1380	21.8812	0.0000	✓ Yes
ENV_SOILMOIS	0.1291	0.0104	0.0181	0.0542	0.3232	6.6268	0.0181	✓ Yes
COVER_RICHNESS_JUN	0.1442	0.0042	0.0081	0.0971	0.1197	8.3044	0.0081	✓ Yes
COVER_TOTAL_JUN	0.1628	0.0012	0.0032	0.1310	0.0343	10.6422	0.0032	✓ Yes
COVER_SHANNON_JUN	0.1117	0.0268	0.0442	0.0739	0.2008	4.9404	0.0442	✓ Yes
COVER_RICHNESS_JUL	0.0419	0.4072	0.4977	0.0408	0.4474	0.6884	0.4977	✗ No
COVER_TOTAL_JUL	0.0337	0.5050	0.5746	0.1353	0.0298	0.4453	0.5746	✓ Yes
COVER_SHANNON_JUL	0.0583	0.2491	0.3288	0.0620	0.2791	1.3323	0.3288	✗ No
COVER_RICHNESS_AUG	0.1616	0.0013	0.0032	0.1809	0.0021	10.4780	0.0032	✓ Yes
COVER_TOTAL_AUG	0.0338	0.5044	0.5746	0.0749	0.2008	0.4465	0.5746	✗ No
COVER_SHANNON_AUG	0.0876	0.0828	0.1301	0.0909	0.1380	3.0248	0.1301	✗ No
PC_SHOOTMASS	0.2348	0.0000	0.0000	0.2260	0.0001	22.8134	0.0000	✓ Yes
PC_LITTERMASS	0.0209	0.6800	0.7238	0.0925	0.1380	0.1704	0.7238	✗ No
PC_HEIGHT	0.1765	0.0004	0.0014	0.2095	0.0003	12.5760	0.0014	✓ Yes
ENV_AIRTEMP	0.0797	0.1146	0.1645	0.0627	0.2791	2.5002	0.1645	✗ No
GH_COUNT	-0.1493	0.0030	0.0066	-0.1456	0.0210	8.9195	0.0066	✓ Yes
PASTURE_TOTALAREA	-0.2354	0.0000	0.0000	-0.2265	0.0001	22.9443	0.0000	✓ Yes
PASTURE_FORESTAREA	0.0745	0.1406	0.1933	-0.0040	0.9372	2.1804	0.1933	✗ No
LAT	0.2170	0.0000	0.0001	0.1140	0.0656	19.3290	0.0001	✓ Yes
LONG	0.2208	0.0000	0.0001	0.2006	0.0005	20.0362	0.0001	✓ Yes
ENV_PROBEMOIS_AUG	0.1784	0.0004	0.0014	0.1142	0.0656	12.8565	0.0014	✓ Yes
ENV_PROBEMOIS_JUL	0.1609	0.0014	0.0032	0.1075	0.0839	10.3931	0.0032	✓ Yes
ENV_PROBEMOIS_JUN	0.0817	0.1059	0.1588	0.0561	0.3171	2.6267	0.1588	✗ No

Download CSV

Feature Importance (Random Forest) Help

Rank	Feature	Importance (Mean)	SD
1	PLOT	0.2053	0.1230
2	COVER_SHANNON_AUG	0.0735	0.0258
3	PC_SHOOTMASS	0.0616	0.0097
4	LONG	0.0547	0.0075
5	COVER_RICHNESS_JUN	0.0537	0.0303
6	PC_HEIGHT	0.0532	0.0051
7	ENV_AIRTEMP	0.0432	0.0056
8	ENV_SOILCLAY	0.0306	0.0112
9	ENV_SOILSILT	0.0297	0.0109
10	COVER_TOTAL_JUN	0.0274	0.0157
11	PASTURE_TOTALAREA	0.0264	0.0038
12	COVER_RICHNESS_JUL	0.0262	0.0054
13	ENV_SOILSAND	0.0256	0.0046
14	ENV_PROBEMOIS_AUG	0.0224	0.0077
15	PC_LITTERMASS	0.0222	0.0033
16	COVER_RICHNESS_AUG	0.0194	0.0133
17	COVER_TOTAL_AUG	0.0172	0.0030
18	COVER_SHANNON_JUN	0.0167	0.0047
19	ENV_SOILNO3	0.0167	0.0048
20	GH_COUNT	0.0164	0.0068
21	COVER_SHANNON_JUL	0.0162	0.0037
22	COVER_TOTAL_JUL	0.0159	0.0066
23	ENV_SOILNH4	0.0159	0.0077
24	ENV_PH	0.0156	0.0054
25	ENV_SOILMOIS	0.0151	0.0047
26	ENV_PROBEMOIS_JUL	0.0144	0.0073
27	ENV_SOILP	0.0134	0.0054
28	ENV_PROBEMOIS_JUN	0.0114	0.0054
29	LAT	0.0098	0.0062
30	ENV_SOILTC	0.0083	0.0011
31	ENV_SOILOC	0.0082	0.0024
32	ENV_SOILTN	0.0073	0.0027
33	PASTURE_FORESTAREA	0.0065	0.0016

Download CSV

Permutation Importance Help

Rank	Feature	Perm. Importance (Mean)	SD
1	PC_SHOOTMASS	0.1094	0.0191
2	LONG	0.0555	0.0237
3	PASTURE_TOTALAREA	0.0305	0.0056
4	ENV_SOILCLAY	0.0289	0.0090
5	COVER_RICHNESS_AUG	0.0275	0.0168
6	COVER_RICHNESS_JUN	0.0209	0.0124
7	PLOT	0.0163	0.0039
8	ENV_SOILMOIS	0.0159	0.0160
9	ENV_PROBEMOIS_AUG	0.0120	0.0098
10	COVER_SHANNON_AUG	0.0119	0.0325
11	PC_HEIGHT	0.0108	0.0153
12	COVER_TOTAL_JUN	0.0101	0.0021
13	ENV_SOILSAND	0.0084	0.0108
14	ENV_SOILNO3	0.0080	0.0061
15	COVER_TOTAL_AUG	0.0074	0.0118
16	COVER_RICHNESS_JUL	0.0073	0.0073
17	COVER_SHANNON_JUN	0.0061	0.0059
18	LAT	0.0051	0.0129
19	ENV_PROBEMOIS_JUL	0.0049	0.0077
20	PASTURE_FORESTAREA	0.0048	0.0033
21	ENV_PH	0.0046	0.0082
22	ENV_SOILNH4	0.0041	0.0027
23	ENV_SOILP	0.0028	0.0034
24	ENV_PROBEMOIS_JUN	0.0027	0.0054
25	ENV_AIRTEMP	0.0021	0.0129
26	PC_LITTERMASS	0.0000	0.0131
27	COVER_SHANNON_JUL	-0.0002	0.0170
28	ENV_SOILTN	-0.0007	0.0039
29	COVER_TOTAL_JUL	-0.0024	0.0038
30	ENV_SOILSILT	-0.0034	0.0060
31	ENV_SOILOC	-0.0053	0.0034
32	ENV_SOILTC	-0.0059	0.0088
33	GH_COUNT	-0.0118	0.0085

Download CSV

SHAP Feature Importance Help

Rank	Feature	SHAP Mean \|Abs\|
1	PC_SHOOTMASS	0.1251
2	LONG	0.1244
3	ENV_SOILCLAY	0.1009
4	PASTURE_TOTALAREA	0.0856
5	PLOT	0.0708
6	PC_HEIGHT	0.0690
7	ENV_SOILSAND	0.0494
8	ENV_SOILSILT	0.0445
9	COVER_SHANNON_AUG	0.0349
10	ENV_AIRTEMP	0.0334
11	ENV_SOILNO3	0.0282
12	GH_COUNT	0.0276
13	COVER_TOTAL_JUN	0.0274
14	ENV_PH	0.0263
15	ENV_PROBEMOIS_AUG	0.0234
16	PC_LITTERMASS	0.0231
17	COVER_RICHNESS_JUN	0.0225
18	ENV_SOILMOIS	0.0212
19	ENV_SOILP	0.0207
20	ENV_SOILNH4	0.0195
21	ENV_PROBEMOIS_JUL	0.0182
22	COVER_RICHNESS_JUL	0.0175
23	LAT	0.0170
24	COVER_RICHNESS_AUG	0.0168
25	COVER_SHANNON_JUN	0.0143
26	COVER_TOTAL_AUG	0.0133
27	COVER_TOTAL_JUL	0.0122
28	COVER_SHANNON_JUL	0.0112
29	PASTURE_FORESTAREA	0.0105
30	ENV_PROBEMOIS_JUN	0.0104
31	ENV_SOILTC	0.0053
32	ENV_SOILOC	0.0050
33	ENV_SOILTN	0.0036

Download CSV

Model Metrics

Split	R²	MSE
Split 1	-0.7322	3.6984
Split 2	0.1112	4.2618
Split 3	0.2379	2.3021
Mean	-0.1277	3.4207

Download CSV

Figures

Feature importance Download PNG

Permutation importance Download PNG

SHAP summary Download PNG

Observed vs predicted (out-of-fold) Download PNG

Cross-validation score distribution Download PNG

Pathway Analysis (same dataset) Help

Run Pathway Analysis

Method & Provenance

Method: Random Forest (Regression)
Method key: random_forest (result type: random_forest, result #149)
Engine: Python · scikit-learn v1.8.0
Dataset(s): #703 gallery — catg_example_08_woody_encroachment.csv
Status: completed
Permanent link: https://cat-g-omxpert.wishartlab.com/results/by-token/c2d3ecce80614f57be4dc0be42e7fcdb Bookmark or share this URL — it does not depend on your session.
Created: 2026-09-21 02:16:12.854253

Parameters used

Parameter	Value
`task_type`	regression
`model_type`	regression
`target_column`	PC_WOODMASS
`n_estimators`	60
`n_splits`	3
`test_size`	0.25
`run_univariate`	True
`run_permutation`	True
`run_shap`	True
`permutation_repeats`	10
`fdr_threshold`	0.05

Execution trace

Reconstructed from the parameters stored with this result — represents the Python/R calls the analysis service made.

from sklearn.ensemble import RandomForestRegressor

model = RandomForestRegressor(n_estimators=60, random_state=42)

scores = sklearn.model_selection.cross_val_score(model, X, y, cv=3)