[Feature Request] Detect redundant relations

[Polfo]

Currently, way 214190421 is part of 4 relations. Relations 3265927 and 6705827 both are part of the same network and carry the same note = 60-298 tag.

Could knooppuntnet provide a warning about this situation. The only current warning is about an incomplete route, while a (different) complete route exists.

[vmarc]

@Polfo thanks for suggesting this new functionality

As a first step, I wrote some code to explore the possibilities for detecting duplicate routes. The code looks for routes with the same name that have at least one way in common. The result is here.

Next would be to figure out if this result actually makes sense. Really duplicate routes? Not too many false positives? The amount of routes found in Germany seems suspiciously high.

[Polfo]

@vmarc Great. Thanks. Interesting. In addition it would be nice to see for each row in the result table: the number of ways and the number of shared ways. Columns could be name, route 1, # ways 1, # shared, # ways 2, route 2

[vmarc]

The report has been updated to include the suggested columns, and a percentage overlap. This makes it easier to distinguish between exact duplicates (100% overlap) and routes with only partial overlap.

The routes with partial overlap have to be looked at more closely to see if it would make sense to issue warnings for those.

For example: there seem to be cases where there are 2 alternative routes between the same nodes. For these it seems to be perfectly ok to have some overlap. It would not be good to issue a warning in these cases. It would be nice if we could somehow recognize these cases by looking at the tags.

Further assistance to analyze the situation is certainly welcome.

[Polfo]

Hello vmarc. Thanks for this quick update. I analysed 25 entries in the table and came to the following logic.

IF ((both routes are OK) and (NL or BE))

    IF (the overlap is 100%) 
        // 19 cases on a total of 25 checks
        One route is redundant
        It can probably be fixed by deleting one route (but check/merge tags)

    IF (the overlap is < 100%) 
        // 5 cases on a total of 25 checks
        It is probably an alternative route (check and suggest alternative route tagging?)

ELSEIF (one route is NOK (missing node))

    IF (the route with less ways is NOK and shares 100% of ways with route with more ways)
        // 1 case on a total of 25 checks
        The small route is probably redundant
        It can probably be fixed by deleting the route (but check parent network is identical).

Note: sometimes, there are > 2 alternative ways like when 4 lock doors (nl:sluis), each representing an alternative. That makes your table longer, but does not change the algorithms. e.g. currently relations 3294691 3294692 3294693 3294694

Aditionally, I looked at 5 German cases. The routes with identical ways and note have different tags, like different name and description. Neuss (1) - Neuss (81) vs KPN Neuss (01) - (81) Neuss KPN Soest (92) - (93) Bad Sassendorf vs RSW (SO-92) Soest-Thöningsen - (SO-93) Bad Sassendorf I assume they generated that programatically and not manually. But at least some redundant routes are maintained in by one user. I don't understand why they do that:

[vmarc]

There are currently 4 routes that are tagged with state=alternate as documented in the wiki. This is shown in 2 additional columns in the report.

[Polfo]

I added encountered frequency in the pseudocode above

[vmarc]

Report 21 may 2024:

nl/hiking 46561 routes, with 43 overlaps

name	route 1	route 2	overlap	# ways 1	# ways 2	# shared	alt 1	alt 2
03-51	6658320	6658321	35%	7	16	4
07-08	3518596	7155614	57%	3	4	2
08-25	11749326	11749327	3%	18	45	1
08-77	1830946	1830952	74%	10	9	7
12-17	4215058	6581626	80%	5	5	4
12-68	11562732	12432648	70%	30	27	20
17-49	6186440	6186441	40%	2	3	1
17-93	5161306	5161307	76%	19	15	13
18-x	7575214	11046624	67%	5	4	3
22-23	11746583	11746584	18%	7	4	1
23-24	7317743	7317744	7%	10	18	1
24-25	11749123	11751730	9%	14	33	2
28-40	11764499	11764500	24%	17	8	3
28-vv	7776398	9174227	50%	2	2	1
29-33	3218910	3218911	18%	4	7	1
32-x	7175609	11047960	71%	7	10	6
34-36	7742677	9638374	40%	2	3	1
36-84	3733997	3734008	50%	2	2	1
39-x	10181051	11047905	50%	3	5	2
40-41	4233340	4587713	44%	4	5	2
44-53	4228903	4613208	57%	7	7	4		ALTERNATE
51-245	15956009	15956010	100%	8	8	8
53-92	1525060	3678746	40%	5	5	2
56-T	4229792	4651524	67%	3	3	2
62-64	3280892	6730084	15%	9	4	1
63-91	6610967	11130821	10%	9	11	1
71-72	3593796	3593797	18%	7	4	1
72-74	5638789	7633251	100%	2	2	2
74-75	11743538	11743539	33%	17	7	4
76-77	11742672	11742673	17%	6	6	1
83-94	2222640	2222641	40%	3	2	1
85-88	10313192	10313193	18%	7	4	1
91-92	3807088	4034136	29%	3	4	1
94-95	7708410	7708411	29%	9	5	2
C07-C09	8383345	8383349	27%	12	3	2
C16-C25	12240946	12240947	40%	2	3	1
E11-E16	7799780	7800949	57%	3	4	2
E63-E94	11906637	11906638	25%	2	6	1
F36-F43	11366571	11366572	18%	7	4	1
L56-L59	13017167	13017168	43%	6	8	3
V32-V33	8026220	9865088	37%	15	12	5
W18-W51	8303295	8303296	60%	6	4	3
W39-W40	12224912	12225776	10%	8	13	1

nl/cycling 15657 routes, with 8 overlaps

name	route 1	route 2	overlap	# ways 1	# ways 2	# shared	alt 1	alt 2
01-87	7322272	7322601	8%	19	30	2
08-10	148820	6334533	31%	7	6	2
22-91	1922281	13226228	25%	7	9	2
33-72	1285470	4757484	18%	6	5	1
39-61	1866196	12896009	32%	10	9	3
44-97	1245517	4769578	38%	9	7	3
53-88	3246843	3308552	15%	19	22	3
66-80	13507259	13507260	6%	16	19	1

nl/horse-riding 2021 routes, with 2 overlaps

name	route 1	route 2	overlap	# ways 1	# ways 2	# shared	alt 1	alt 2
20-92	3737991	3737992	30%	9	11	3
61-96	8484376	17370984	67%	1	2	1

be/hiking 15439 routes, with 21 overlaps

name	route 1	route 2	overlap	# ways 1	# ways 2	# shared	alt 1	alt 2
02-99	12812702	12812916	15%	6	7	1
03-90	6856566	6934504	22%	2	7	1
05-06	2336935	2336936	11%	6	13	1
05-06	7768744	7768770	55%	6	5	3
09-85	2393578	17460083	100%	2	2	2
11-56	2569046	2569052	40%	3	2	1
13-14	2851007	7421410	67%	4	5	3
17-18	3294691	3294692	77%	13	13	10
17-18	3294691	3294693	26%	13	10	3
17-18	3294691	3294694	25%	13	11	3
17-18	3294692	3294693	26%	13	10	3
17-18	3294692	3294694	25%	13	11	3
17-18	3294693	3294694	67%	10	11	7
30-92	4187259	4187260	44%	3	6	2
34-35	2569065	2569066	44%	4	5	2
45-60	2171775	2171776	31%	6	7	2
48-49	308368	308553	20%	3	7	1
51-56	6085464	7368966	50%	2	2	1
55-57	7768539	7768540	11%	7	11	1
55-69	12695702	12695703	25%	4	4	1
58-59	2669998	15447335	100%	4	4	4

be/cycling 10433 routes, with 17 overlaps

name	route 1	route 2	overlap	# ways 1	# ways 2	# shared	alt 1	alt 2
02-87	12381947	12952747	100%	6	6	6
05-82	11779746	12982487	100%	14	14	14
06-14	7340458	7340476	34%	25	28	9		ALTERNATE
26-30	11771707	12982120	84%	31	31	26
26-47	7443274	16705027	51%	20	15	9
31-34	12357	1764164	53%	20	18	10
50-51	967549	12385292	100%	10	10	10
51-54	1840378	12385306	100%	16	16	16
52-72	12380217	12952767	61%	15	8	7
53-54	9250	3935684	80%	3	2	2
55-56	11795955	12988981	83%	5	7	5
61-71	1843936	11779728	82%	35	33	28
72-73	11774117	12982172	95%	10	11	10
75-182	1772965	15659173	76%	26	32	22
83-92	11781768	12946850	100%	20	20	20
92-93	11779732	12937172	100%	10	10	10
92-95	19005	7151646	36%	17	16	6	ALTERNATE

de/hiking 13745 routes, with 14 overlaps

name	route 1	route 2	overlap	# ways 1	# ways 2	# shared	alt 1	alt 2
21-23	8488396	8488400	14%	3	11	1
22-23	8488397	8488399	33%	8	10	3
90-93	12844769	14438924	50%	1	3	1
Am Hochmoor - Mooshof	6130257	6130258	22%	5	4	1
Am Klingelberg - Stollenwald-Hölzle	11973421	11973422	33%	3	3	1
Am Rumpelstein - Wingolfbrunnen	13048482	13048483	18%	4	7	1
Kiechlinsbergen Beim Käpelle - ?	17270915	17270916	40%	3	2	1
Mittlerer Busterbach - Oberes Silbergründle	12096650	12096651	36%	6	5	2
Muckenhofweiher - Renzenhof	14400600	14400601	17%	7	5	1
Obere Sonnenhalde - Sonnenhalde	17148579	17148583	40%	3	2	1
Riesenbühlweg - Vogelhaus	14276356	14276363	25%	6	2	1
Schanzenbergsattel - Schloss Bad Rotenfels	16755253	16755254	17%	10	2	1
Seibelseckle - Skilift Darmstädter Hütte	12731869	12731871	13%	5	10	1
Theurerswald - Wanderheim Zavelstein	10041108	17191869	33%	4	2	1

de/cycling 13042 routes, with 31 overlaps

name	route 1	route 2	overlap	# ways 1	# ways 2	# shared	alt 1	alt 2
01-48	13254981	13255003	50%	8	4	3
04-05	5941416	13948342	100%	22	22	22
04-05	12892395	13118633	100%	9	9	9
17-79	7711579	12365526	100%	3	3	3
20-69	7711537	12365558	100%	3	3	3
21-37	7706766	9240482	100%	32	32	32
22-27	5391806	5391807	44%	10	8	4
22-27	5391806	7820144	44%	10	8	4
22-27	5391807	7820144	100%	8	8	8
22-37	7752176	9240493	100%	3	3	3
29-31	9227315	13992715	100%	5	5	5
29-32	7908654	7908655	71%	27	38	23
29-32	15959339	15959871	5%	21	21	1		ALTERNATE
31-46	123734	2247758	44%	8	10	4
39-94	12936079	15088682	20%	1	9	1
44-46	1829105	2294733	6%	8	23	1
45-46	6009045	7721837	16%	94	32	10
56-72	7592204	7701811	100%	24	24	24
59-90	9790634	12381299	100%	3	3	3
61-68	14481177	15092450	80%	2	3	2
69-70	5997475	7736699	4%	32	22	1
69-71	1830736	2369015	25%	33	32	8
69-81	11928453	12346940	100%	1	1	1
70-71	2271346	2369135	20%	24	27	5
71-72	2271986	2369897	27%	14	30	6
71-99	9790632	12429000	100%	1	1	1
80-82	7997925	8041876	22%	10	26	4
81-83	7528368	7542584	11%	44	12	3
83-85	8917224	8917225	11%	11	8	1
92-94	9796376	12381292	100%	10	10	10
Sch21 - Sch21	12966407	12966409	75%	14	10	9

fr/hiking 20420 routes, with 56 overlaps

name	route 1	route 2	overlap	# ways 1	# ways 2	# shared	alt 1	alt 2
? - Col de la Chapelle	12356391	12356392	25%	3	5	1
? - Doucy Village	15044922	15082036	44%	6	3	2
? - Giron	13672670	13672671	20%	5	5	1
? - La Borne au Lion	13644061	13672683	14%	12	2	1
? - Le Chaleyras	17611236	17611240	67%	2	1	1
? - Le lac de Pramol	14874322	14874324	22%	3	6	1
? - Les Côtes	12120715	15966177	25%	4	4	1
? - Montagne d'en Haut	15068920	15068923	33%	3	3	1
? - Refuge des Évettes	15033347	15033348	40%	2	3	1
? - Saint-Geney	17537082	17538631	29%	4	3	1
?-?	13473679	14156925	12%	4	13	1
?-?	13569606	13573807	25%	4	4	1
?-?	13659816	15430260	30%	7	26	5
?-?	13669140	13669142	25%	3	5	1
?-?	13669143	13669144	17%	9	3	1
?-?	13669147	13669149	25%	6	2	1
?-?	13825759	13825761	57%	3	4	2
?-?	14011898	14020559	50%	2	2	1
Auberives-en-Royans - Le Barrage	17532880	17556476	100%	21	21	21
Baraque de Carpe - Gampaloux	17508473	17508476	67%	4	5	3
Belvédère des Expertois - Chaumois Boivin	13765922	13765923	33%	2	4	1
Bois Sigü - Ecole d'escalade	12432286	17471222	100%	1	1	1
Cable Keller - Le Châtaignier	14137618	14954082	100%	1	1	1
Champlan - ?	12531090	12531109	50%	2	2	1
Chemin d'Herbe - Pas des Feuillées	17497195	17569388	89%	9	9	8
Chichilianne - Chichilianne	17498125	17501504	100%	3	3	3
Col de Papavet - Les Sages	17498122	17498132	50%	2	2	1
Col des Pellas - Col de Papavet	17498679	17498680	50%	2	2	1
Doucy Village - Oratoire du Chaussis	15044924	15082037	40%	3	2	1
Ecole d'escalade - La fontaine du val Furon	12432287	17471221	100%	4	4	4
Grande Aiguille - Lac des Pourettes	2518488	14512583	100%	1	1	1
L'Igloo - Croix des 7 Frères	17428301	17428561	100%	2	2	2
La Basse Combe Froide - Sur les Puits	13661460	13661461	67%	3	3	2
La Combe Nord - Bourbourin	16061231	16387457	60%	5	5	3	ALTERNATE
La Frette - Cable Keller	14185528	14954083	100%	4	4	4
La Piscine - ?	17462266	17462267	17%	3	9	1
La fontaine du val Furon - Bois Sigü	12432288	17471220	100%	3	3	3
Le Châtaignier - Font Chaise	14137617	14954081	86%	7	7	6
Le Moulin - ?	17507243	17507244	17%	5	7	1
Le Solliet (Sud) - Le Solliet (Nord)	17434111	17434112	100%	1	1	1
Les Carrières - La Planche du Fenu	14188445	14188446	25%	4	4	1
Les Charbonnières - Roches de Saint-Béron	14459911	14749452	67%	1	2	1
Les Civières - Gève	12434636	17445917	20%	5	5	1
Les Poules - Pré Achard	17479167	17520073	100%	1	1	1
Les Sages - ?	17498121	17498133	50%	2	2	1
Mas Rebuffat - Le Passeït	17556232	17611224	100%	3	3	3
Mollard des Barres - ?	13573799	13574532	20%	7	3	1
Moure de Trian - Chêne Rond	17047141	17279311	33%	3	3	1
Oppedette (Mairie) - Parking des Belvédères	16020492	16020493	80%	10	10	8
Passerelle du Foron - Lac de Flérier	17408940	17409019	33%	2	4	1
Plan Corbet - Col de la Légette	17195212	17195213	50%	2	2	1
Rue des Ondines - Pont de la Poudrerie	12232602	12346014	62%	7	6	4		ALTERNATE
Saint-Nazaire-en-Royans - ?	17526240	17526241	17%	4	8	1
Salastre - Notre Dame d'Ajude	14769975	15880994	29%	3	4	1
Tiers de Châtillon - ?	17410458	17410497	67%	1	2	1
Écomusée de l'Ocre - Piquebori	16363407	16363419	59%	8	9	5		ALTERNATE

fr/cycling 1792 routes, with 7 overlaps

name	route 1	route 2	overlap	# ways 1	# ways 2	# shared	alt 1	alt 2
05-15	15587866	15617862	100%	7	7	7
65-67	15576803	15576810	100%	1	1	1
66-68	15572218	15582738	86%	12	16	12
81-77	15773335	16836114	100%	12	12	12
84-86	15668726	15671392	100%	4	4	4
91-90	16266338	16841944	100%	2	2	2
94-41	16841939	17163753	100%	3	3	3

at/cycling 132 routes, with 1 overlaps

name	route 1	route 2	overlap	# ways 1	# ways 2	# shared	alt 1	alt 2
32-34	10440632	10440633	15%	9	4	1	ALTERNATE

es/hiking 275 routes, with 1 overlaps

name	route 1	route 2	overlap	# ways 1	# ways 2	# shared	alt 1	alt 2
AE216 - AE220	11351581	11351589	5%	25	18	1

[pyrog]

related to https://github.com/JOSM/pt_assistant/issues/66