P seq spikes
Kate
Christopher Ramsey
I think given the spikes are only in the lower elements it is probably the very unconstrained solutions which are causing the issue (low k). Given this the first think to try might be to raise the lower limit on this with U(-1,2) or U(-0.5,2) - you can check the marginal posterior for this k distribution to see what is going on.
Best wishes
Christopher
> You received this message because you are subscribed to the Google Groups "OxCal" group.
> To unsubscribe from this group and stop receiving emails from it, send an email to oxcal+un...@googlegroups.com.
> To view this discussion visit https://groups.google.com/d/msgid/oxcal/3e5cb3aa-f724-4a2d-aa9a-1cc2790f8892n%40googlegroups.com.
> <Screenshot 2025-03-24 at 3.51.53 PM.png>
Kate
Christopher Ramsey
Ok- unless you want to post the full code here you could contact me directly to see if I can see why this is happening - it looks as if you have poor covergence for some reason. The plot suggests that you have some internal structure in the sequence though which I don't quite understand - as you should not have Phases in a P_Sequence?
Best wishes
Christopher
> <Screenshot 2025-03-27 at 10.17.46 AM.png><Screenshot 2025-03-27 at 10.11.07 AM.png>
Kate
Plot()
{
P_Sequence("PCR4c", 0.1, 1, U(-0.5,2))
{
Boundary("Bottom")
{
z=620;
};
R_Date("26.1-QL-1967",1410.0,14.1)
{
z=618;
};
R_Date("A-2150",1524,112)
{
z=609.5;
};
R_Date("A-2153",1519,88)
{
z=609.5;
};
R_Date("USGS-902",1351,56)
{
z=599.5;
};
R_Date("QL-1969",1323.0,16.2)
{
z=599.5;
};
Date("EQ C")
{
z=597.5;
};
R_Date("QL-1968",1270,15.8)
{
z=597.5;
};
R_Date("31-USGS-901",1336,80)
{
z=579.5;
};
R_Date("32.1-186855",1265,35)
{
z=554.5;
};
R_Date("186853",1285,30)
{
z=491.5;
};
R_Date("33.1-143588",1280,30)
{
z=456.5;
};
R_Date("143590",1300,30)
{
z=450.5;
};
R_Date("QL-1970",1259.0,15.7)
{
z=450.5;
};
R_Date("QL-1971",1291.0,15.8)
{
z=444.5;
};
R_Date("143591",1280,30)
{
z=444.5;
};
R_Date("34.1-143531",1250,30)
{
z=436.5;
};
Date("EQ D")
{
z=430;
};
R_Date("35-143593",1260,30)
{
z=376.6;
};
R_Date("186781",1175,30)
{
z=373.5;
};
R_Date("136467",1140,55)
{
z=373.5;
};
R_Date("QL-1950",1223.0,15.8)
{
z=360.5;
};
R_Date("QL-1973",1215.0,16.7)
{
z=360.5;
};
R_Date("136463",1190,30)
{
z=360.5;
};
R_Date("140077",1185,35)
{
z=360.5;
};
R_Date("I-9590",1052,143)
{
z=360.5;
};
Date("EQ F")
{
z=359;
};
R_Date("41-USGS-139",1162,111)
{
z=338.5;
};
R_Date("43.1-136487",1090,60)
{
z=326;
};
R_Date("136489",1165,60)
{
z=312;
};
R_Date("136488",1110,60)
{
z=312;
};
R_Date("136473",1115,60)
{
z=310;
};
R_Date("143529",1085,30)
{
z=309.5;
};
R_Date("186802",1160,40)
{
z=306.5;
};
R_Date("USGS-138",1103,119)
{
z=305.5;
};
R_Date("QL-1951",1076.1,17.1)
{
z=305.5;
};
Date("EQ I")
{
z=303;
};
R_Date("QL-1952",1058.3,15.7)
{
z=301.5;
};
R_Date("QL-1976",1005,16.2)
{
z=301.5;
};
R_Date("136474",915,70)
{
z=301.5;
};
R_Date("188717",1085,40)
{
z=300.5;
};
R_Date("188641",1080,40)
{
z=300.5;
};
R_Date("143536",1090,35)
{
z=299.5;
};
R_Date("188640",1070,35)
{
z=299.5;
};
R_Date("USGS-83",1107,87)
{
z=292.5;
};
R_Date("136455",875,35)
{
z=292.5;
};
R_Date("136456",835,60)
{
z=292.5;
};
R_Date("52-140076",950,30)
{
z=262.5;
};
Date("EQ N")
{
z=255;
};
R_Date("UW-573",907,95)
{
z=219.5;
};
R_Date("55-143594",970,30)
{
z=211.5;
};
R_Date("59.2-143597",910,30)
{
z=196.5;
};
Date("EQ R")
{
z=196.5;
};
R_Date("QL-1977b",906.0,13.6)
{
z=196.5;
};
R_Date("QL-1977a",866.0,16.0)
{
z=196.5;
};
R_Date("QL-1977c",854.0,15.5)
{
z=196.5;
};
R_Date("59.4-143595",905,30)
{
z=174.5;
};
R_Date("147688",840,30)
{
z=169.5;
};
R_Date("147689",830,25)
{
z=169.5;
};
R_Date("147687",820,25)
{
z=169.5;
};
R_Date("QL-1955",816.3,14.6)
{
z=169.5;
};
R_Date("188644",815,30)
{
z=169.5;
};
R_Date("QL-1978",814.0,15.4)
{
z=169.5;
};
R_Date("USGS-84",783,111)
{
z=168.5;
};
R_Date("UW-365",663,71)
{
z=166.5;
};
R_Date("USGS-896",624,56)
{
z=166.5;
};
R_Date("188645",715,35)
{
z=165.5;
};
R_Date("186778",680,30)
{
z=165.5;
};
R_Date("114089",640,35)
{
z=165.5;
};
R_Date("140075",615,30)
{
z=165.5;
};
R_Date("QL-1953",601.4,16.9)
{
z=165.5;
};
R_Date("QL-1954",572,12.8)
{
z=165.5;
};
R_Date("QL-1979",571.0,15.2)
{
z=165.5;
};
R_Date("113998",515,40)
{
z=165.5;
};
Date("EQ T")
{
z=164;
};
R_Date("68.1-QL-1995",542,15.0)
{
z=153.5;
};
R_Date("186796",425,35)
{
z=151.5;
};
R_Date("186776",405,30)
{
z=151.5;
};
R_Date("186797",375,35)
{
z=151.5;
};
R_Date("186777",375,25)
{
z=151.5;
};
R_Date("186775",360,25)
{
z=151.5;
};
R_Date("186773",375,25)
{
z=147.5;
};
R_Date("186772",375,25)
{
z=147.5;
};
R_Date("186771",375,25)
{
z=147.5;
};
R_Date("186774",355,25)
{
z=147.5;
};
R_Date("USGS-895",302,64)
{
z=147.5;
};
R_Date("136491",370,35)
{
z=145.5;
};
R_Date("113996",355,35)
{
z=145.5;
};
R_Date("QL1956",344.5,16.6)
{
z=145.5;
};
R_Date("QL-1957",342.0,16.5)
{
z=145.5;
};
R_Date("113997",285,35)
{
z=145.5;
};
Date("EQ V")
{
z=143;
};
R_Date("113995",350,35)
{
z=142.5;
};
R_Date("143533",290,30)
{
z=142.5;
};
R_Date("USGS-136",437,111)
{
z=140.5;
};
R_Date("75-QL-1992",260.0,16.3)
{
z=125.5;
};
R_Date("77-143596",220,30)
{
z=103.5;
};
R_Date("114091",195,35)
{
z=78.5;
};
R_Date("QL-1991",165.0,16.1)
{
z=78.5;
};
R_Date("USGS-144",163,95)
{
z=78.5;
};
R_Date("UW-347",83,143)
{
z=78.5;
};
R_Date("114000",220,35)
{
z=67.5;
};
R_Date("114090",210,50)
{
z=67.5;
};
R_Date("QL-1981",183.0,9.3)
{
z=67.5;
};
R_Date("QL-1980",171.5,12.8)
{
z=67.5;
};
R_Date("114001",105,40)
{
z=67.5;
};
Date("EQ X")
{
z=66;
};
R_Date("88-QL-1990",93.0,16.2)
{
z=47.5;
};
Boundary("1857", 1857)
{
z=46;
};
};
};
Christopher Ramsey
Ok - the main issue here is that the P_Sequence (like the Sequence) implies the events are in order - this means that you cannot have two at the same depth as they would be at the same time. So you have to use R_Combine where there are multiple dates. For the end boundary it also helps if there is at least some uncertainty. The following code should do what you want I think:
Plot()
{
P_Sequence("PCR4c", 0.1, 1, U(-2,2))
{
{
z=620;
};
R_Date("26.1-QL-1967",1410.0,14.1)
{
z=618;
};
{
R_Date("A-2150",1524,112);
R_Date("A-2153",1519,88);
z=609.5;
};
R_Combine()
{
R_Date("USGS-902",1351,56);
R_Date("QL-1969",1323.0,16.2);
z=599.5;
};
Date("EQ C",R_Date("QL-1968",1270,15.8))
{
};
R_Date("31-USGS-901",1336,80)
{
z=579.5;
};
R_Date("32.1-186855",1265,35)
{
z=554.5;
};
R_Date("186853",1285,30)
{
z=491.5;
};
R_Date("33.1-143588",1280,30)
{
z=456.5;
};
{
R_Date("143590",1300,30);
R_Date("QL-1970",1259.0,15.7);
z=450.5;
};
R_Combine()
{
R_Date("QL-1971",1291.0,15.8);
z=444.5;
};
R_Date("34.1-143531",1250,30)
{
z=436.5;
};
Date("EQ D")
{
z=430;
};
R_Date("35-143593",1260,30)
{
z=376.6;
};
{
R_Date("186781",1175,30);
R_Date("136467",1140,55);
z=373.5;
};
R_Combine()
{
R_Date("QL-1950",1223.0,15.8);
R_Date("QL-1973",1215.0,16.7);
R_Date("136463",1190,30);
R_Date("140077",1185,35);
z=360.5;
};
Date("EQ F")
{
z=359;
};
R_Date("41-USGS-139",1162,111)
{
z=338.5;
};
R_Date("43.1-136487",1090,60)
{
z=326;
};
{
R_Date("136489",1165,60);
z=312;
};
R_Date("136473",1115,60)
{
z=310;
};
R_Date("143529",1085,30)
{
z=309.5;
};
R_Date("186802",1160,40)
{
z=306.5;
};
{
R_Date("USGS-138",1103,119);
z=305.5;
};
Date("EQ I")
{
z=303;
};
{
R_Date("QL-1952",1058.3,15.7);
R_Date("QL-1976",1005,16.2);
R_Date("136474",915,70);
z=301.5;
};
R_Combine()
{
R_Date("188717",1085,40);
R_Date("188641",1080,40);
z=300.5;
};
R_Combine()
{
R_Date("143536",1090,35);
R_Date("188640",1070,35);
z=299.5;
};
R_Combine()
{
R_Date("USGS-83",1107,87);
R_Date("136455",875,35);
z=292.5;
};
R_Date("52-140076",950,30)
{
z=262.5;
};
Date("EQ N")
{
z=255;
};
R_Date("UW-573",907,95)
{
z=219.5;
};
R_Date("55-143594",970,30)
{
z=211.5;
};
{
R_Date("59.2-143597",910,30);
R_Date("QL-1977b",906.0,13.6);
R_Date("QL-1977a",866.0,16.0);
z=196.5;
};
R_Date("59.4-143595",905,30)
{
z=174.5;
};
{
R_Date("147688",840,30);
R_Date("147689",830,25);
R_Date("147687",820,25);
R_Date("QL-1955",816.3,14.6);
R_Date("188644",815,30);
z=169.5;
};
R_Date("USGS-84",783,111)
{
z=168.5;
};
{
R_Date("UW-365",663,71);
R_Date("USGS-896",624,56);
R_Date("188645",715,35);
R_Date("186778",680,30);
R_Date("114089",640,35);
R_Date("140075",615,30);
R_Date("QL-1953",601.4,16.9);
R_Date("QL-1954",572,12.8);
R_Date("QL-1979",571.0,15.2);
z=165.5;
};
Date("EQ T")
{
z=164;
};
R_Date("68.1-QL-1995",542,15.0)
{
z=153.5;
};
{
R_Date("186796",425,35);
R_Date("186776",405,30);
R_Date("186797",375,35);
R_Date("186777",375,25);
R_Date("186775",360,25);
z=151.5;
};
R_Combine()
{
R_Date("186773",375,25);
R_Date("186772",375,25);
R_Date("186771",375,25);
R_Date("186774",355,25);
R_Date("USGS-895",302,64);
z=147.5;
};
R_Combine()
{
R_Date("136491",370,35);
R_Date("113996",355,35);
R_Date("QL1956",344.5,16.6);
R_Date("QL-1957",342.0,16.5);
z=145.5;
};
Date("EQ V")
{
z=143;
};
{
R_Date("113995",350,35);
z=142.5;
};
R_Date("USGS-136",437,111)
{
z=140.5;
};
R_Date("75-QL-1992",260.0,16.3)
{
z=125.5;
};
R_Date("77-143596",220,30)
{
z=103.5;
};
{
R_Date("114091",195,35);
R_Date("QL-1991",165.0,16.1);
R_Date("USGS-144",163,95);
R_Date("UW-347",83,143);
z=78.5;
};
R_Combine()
{
R_Date("114000",220,35);
R_Date("114090",210,50);
R_Date("QL-1981",183.0,9.3);
R_Date("QL-1980",171.5,12.8);
z=67.5;
};
Date("EQ X")
{
z=66;
};
R_Date("88-QL-1990",93.0,16.2)
{
z=47.5;
};
{
z=46;
};
};
};
Best wishes
Christopher
> I had forgotten no phases in P_Seqs! Thanks for that reminder. I removed, and also reconsidered my choice in k. The depth values have fractions (eg 100.5 cm), so I understand k=0.1 would be more appropriate (or converting all depths to mm and using k=1000). I ran both using the stable k (it runs a lot quicker) and this did reduce the spikiness, but some strange splits in the posteriors remain. The variable k approach, P_Sequence("", 0.1, 1, U(-0.5,2)) was problematic, similar to the previous (k distro below). <Screenshot 2025-04-01 at 10.22.09 AM.png>I note that there are a handful of high precision dates with very low uncertainties (~15 yrs)- notably one at the very bottom- and wonder if that sets up the split?