Medical pedigrees with TeX and PSTricks
This is a modal window.
The media could not be loaded, either because the server or network failed or because the format is not supported.
Formal Metadata
| Title |
| |
| Subtitle |
| |
| Title of Series | ||
| Part Number | 22 | |
| Number of Parts | 33 | |
| Author | ||
| License | CC Attribution 3.0 Unported: You are free to use, adapt and copy, distribute and transmit the work or content in adapted or unchanged form for any legal purpose as long as the work is attributed to the author in the manner specified by the author or licensor. | |
| Identifiers | 10.5446/30788 (DOI) | |
| Publisher | ||
| Release Date | ||
| Language | ||
| Production Place | Cork, Ireland |
Content Metadata
| Subject Area | ||
| Genre | ||
| Abstract |
|
1
2
4
7
8
12
17
18
19
20
24
28
30
00:00
InformationTheory of relativityCircleGenetic programmingSquare numberNetwork topologyStandard deviationComputational complexity theoryFormal languageStudent's t-testComputer programmingRevision controlLine (geometry)MereologyTap (transformer)AlgorithmDatabaseCollaborationismMultiplication signUniform resource locatorFigurate numberType theoryIncidence algebraShape (magazine)Symmetric matrixAuthorizationMathematicsRule of inferenceSquaring the circleField (computer science)Diffuser (automotive)View (database)Standard errorPhysical lawArithmetic meanData managementGroup actionComplex (psychology)Volume (thermodynamics)HypermediaRight angleDegree (graph theory)Wage labourComputer fileLink (knot theory)TheoryKettenkomplexHuman migrationFamilySpreadsheetSystem programmingState of matterTouchscreenMachine visionAreaForm (programming)Open sourceSemiconductor memoryNumberEngineering drawing
06:48
Figurate numberMachine codeSystem callTheory of relativityVideo gameGenetic programmingNumberForm (programming)Formal languageComputer programmingLine (geometry)Connected spaceInferenceInsertion lossDatabaseRight angleData managementResultantMereologyScaling (geometry)Programmer (hardware)Degree (graph theory)View (database)Graph (mathematics)Point (geometry)Forcing (mathematics)Open setWage labourComputer fileBitoutputSpreadsheetData dictionaryRoboticsStatistical hypothesis testingService (economics)Endliche ModelltheorieDiagramMarginal distributionFlow separationNormal (geometry)Level (video gaming)Instance (computer science)Term (mathematics)Macro (computer science)Writing2 (number)Logical constantArithmetic meanFamilyTap (transformer)Military baseRow (database)Inheritance (object-oriented programming)WaveTwin primeType theoryComputer fontFile formatComputer animation
13:35
Computer programmingCASE <Informatik>Revision controlCartesian coordinate systemResultantFamilyBitAlgorithmMathematicianVideo gameDegree (graph theory)Point (geometry)ForestView (database)Right angleMereologyArithmetic meanNatural numberMetropolitan area networkLevel (video gaming)Endliche ModelltheorieWordData managementWeightStapeldateiLine (geometry)Digital electronicsComputer fileClosed setArithmetic progressionDistanceSlide ruleLogical constantComputer programmingComputer clusterCuboidNetwork topologyRecursionRootImage resolutionTwo-dimensional space2 (number)RoutingComplex (psychology)Engineering drawing
20:23
Water vaporDegree (graph theory)Execution unitDifferent (Kate Ryan album)Rule of inferenceInstance (computer science)Genetic programmingFamilyCASE <Informatik>Ideal (ethics)NumberBitMultiplication signRevision controlOrder (biology)Reading (process)Computer animationXMLProgram flowchart
26:03
Goodness of fitMarginal distributionEndliche ModelltheorieQuicksortComputer configurationWordLattice (order)Process (computing)Network topologyMoment (mathematics)Ocean currentGenetic programmingKey (cryptography)NumberRight angleShape (magazine)Formal languageMereologyVotingSystem programmingFamilyComputer fileType theoryComputer animation
31:34
MereologyParticle systemGraph (mathematics)Arithmetic meanLine (geometry)Graph (mathematics)Machine visionCategory of beingElectric generatorCASE <Informatik>Projective planeSlide ruleEndliche ModelltheorieData managementoutputComputer programmingSystem programmingCollaborationismSource codeEuler anglesMultiplication signMoment (mathematics)InformationReal numberNichtlineares GleichungssystemVariable (mathematics)Control flowStatistical hypothesis testingDegree (graph theory)Computer programmingError messageLaser scanningForm (programming)Shift operatorDiagramProper mapSemiconductor memoryInheritance (object-oriented programming)Point (geometry)Complex (psychology)Revision controlUsabilityMusical ensembleArrow of timeProbability density functionComputer fileTerm (mathematics)GodNetwork topologyLogic programmingLatent heatWeb pageLogical constantStandard deviationComputer animation
Transcript: English(auto-generated)
00:00
Okay, to say the truth, this talk presented a challenge for me, and mainly because we discussed the previous version of this program two years ago at the PRACtech conference, and it even got into the graphics companion, thank you very much, Frank, by the way,
00:24
for this. So, if I would start from the beginning and just repeat, those of you who know about it will probably be bored. But if I would assume that everybody knows about the previous version and just discuss
00:40
what new is there, I probably would bore the other part of the audience. So what I will do, I will briefly describe what we are doing. I will not talk much about the algorithms, it was a lot of algorithmic work which went into this.
01:01
I will just say, okay, read our fine manual, and you will find there are as much about algorithms as you want. And what I will do, I will just tell what the new things we can do and the new possibilities, and then, since we have time, I will have some small demonstration.
01:22
And by the way, for the demonstration, I will need a volunteer. So I will give you time to think whether you want to volunteer or not. Okay, let's start from the beginning. If you ever come to a doctor who works with hereditary diseases, one of the things
01:41
the doctor would do with you, he or she would ask a lot of rather noisy questions about your parents, about your kids, about your brothers, about your sisters, about nieces, nephews, aunts, everybody.
02:01
I asked my collaborator Leila Ahmadieva how deep is the interview, how many relatives you really should talk about, and her answer was everybody, the patient can remember, and if you can interview the relatives about somebody he doesn't remember, but they remember,
02:24
it's even better. The more is the better. So if you have a chance to know your genealogy up to the 13th century, you would make your doctor very happy. And they write all this down, and let me show what they do with this information.
02:43
They do a picture like this. Actually I think I wrote, I said that this thing is wrong, I said it's a complex pedigree, but this is what I did when I started to work with this. Now I understand that this is not a complex, this is a pretty simple pedigree. It's basically a picture where they put all the information about your relatives
03:03
on paper, and it has a certain formal, highly formal language. For example, the squares are male, the circles are female, they might be filled, they have a lot of thoughts, it depends on their status, whether they were healthy or had some disease
03:25
that the doctor wants about this, want about to know, there are some symbols for abortions, there are some symbols for infertile marriage, and a lot and lot of other things.
03:43
Actually here is a reference to the standard notation, and it's much smaller than the XML standard, but it's as boring as this. And if you think about this, it looks like genealogy tree, but it's much more complex,
04:02
because genealogy tree is a tree, you have a root, you go down, down, and here you have cycles or loops, which makes it very difficult, you have several loops, and I did not do this just for a conference in Ireland, but Otype, obviously an Irish surname, it was in the paper,
04:25
in this paper, so it's a coincidence. Okay, usually people just take the information from the interview and manually put all the circles, squares, and lines on paper, and it takes a lot of time, and a lot of effort,
04:45
and a lot of dropsmanship, so you really would like to do this in some mathematical way, especially if you want to publish it. If you look at the pedigrees that are published in medical journals and medical books,
05:03
some of them are really well done, some of them are less well done, and there are some old school doctors who know it very well, who are well trained, some of the newer generation, well Leila told me that she actually makes her students to do it well, but I think she is in minority here, so we wanted to make a program that would
05:26
do this, that would make pictures like this automatically, and from the beginning we understood that it's very difficult to do automatically, because the algorithmic complexity of making tree-like, okay, it's very easy to make a tree, unfortunately we don't have
05:44
a tree there, so we said, okay, let's do it in the two-way approach, much like the approach that Manoush had told in the first lecture today. First, we have some information about relatives, I called it database, it could be ASCII file,
06:04
unfortunately for most physicians it's actually an Excel spreadsheet, so I will show you that we are unfortunately Excel oriented, you make this database, and from this database you can manually make a TAP file, so you can make a TAP histories package with Macros,
06:27
which says, okay, here is a person, place this person here, here is another person, place this person here, place all the lines for siblings, for aunts, nephews and so on.
06:40
You can make a TAP file, you can process this TAP and PS tricks, and you'll get a nice postscript of PDF, which you can publish, you can look, and you can study. And since it's a rather boring process, it's much less boring than doing it with the parents and paper, but still boring, you'd like to make a program, I put Perl here,
07:06
and with this Perl program it would read your data and make a TAP file. And it's very difficult, it's a usual 80-90 situation, it's very easy to make 80% of work, it's very easy to write a program which will make 80% of p-degrees
07:23
right, it's very difficult to make the remaining 20% p-degrees right. So instead of putting effort here, we say, okay, let's make 80% right, and for the remaining 20% we'll make some adjustment. Somebody will look at the TAP file, make adjustment, and get it done.
07:46
Okay, so first a little bit about TAP part. Well, TAP part is PSTricks based. If you know PSTricks, we have a workshop on Sunday, we have had a talk, first talk today,
08:02
basically the idea you have nodes, and you have connection. So from the point of view of mathematician, your p-degree is a graph, you have nodes on the graph, you have connection, and you need macros to put node, and you need macros to put connection. Nodes are persons, they are abortions,
08:21
they are special symbols for infertile marriages and so on. So you have a macro for person, you can say male to female, adopted, affected, which means had some disease, and a lot of other things. Here you see, for example, this is a symbol for a pregnant female.
08:40
This is a symbol for a male who died at his 20th year, and so on. And for relationship and descent, we use lines, which are node connection on the PSTricks language. So here we have just a pair of male and female.
09:04
Here we have a pair of male and female who divorced. And this is constant meaning relation, which is a relation, which is a marriage between relatives, like cousins or second cousins or whatever.
09:24
And here are a couple, and here are some examples. For example, here is a PSTricks code, and those who work with PSTricks would recognize how it is done. We put a person, another person, we have put some relation,
09:42
and here you have a female which has one adopted daughter, one natural son, and had some miscarriage at some point of her life. He is a marriage which did not have children due to anospermia. Here is a more complex diagram.
10:03
You have a marriage between Fred and Ginger, which had John, Mary, and again, a male miscarriage. Okay, this line is both. This is a symbol which says that this person
10:21
is asymptomatic patient. He has some hereditary disease, but you cannot see it from the symptoms. You can just analyze and find that it's too big to be diagnosed. I see that Leila wants to correct me. Okay, it goes like this.
10:41
Here is a situation when we have twins. We, the two boys, are monozygotic twins, or identical twins in our layman terms. And again, we have nice macros to say this. And this male has four daughters which are twins.
11:02
Sometimes it happens. As you see, there are macros, and if you know any technician, he would be happy to write macros for you. Now let's talk about the wrong part. Again, I don't, I expect some physicians
11:20
and some genetic researchers to be able to write their prose like this. I don't expect all of them to do. So what could they write? First, this is a format of input data. I hope that the font is not too small for you to understand.
11:40
Normally, again, it would be an Excel spreadsheet. I don't have Excel on this Linux desktop, so this is ASCII form of the spreadsheet. As you see, you have columns. Each column says something about, the line is a person. You say it's male or female, date of birth, date of death.
12:01
Mother and fathers are actual names of the lines. Excel can do it automatically for you if you just, in Excel you would just say, okay, he is the mother, he is the father. In ASCII text, you just need to type in the ID, the number of row for this.
12:22
I am certain that Jonton Q can write for you a nice program which would do all this automatically. You would just be able to hand wave and so on. The number of columns could be different. The program is smart enough to understand what's going on, and it looks at this
12:43
and makes something like this. So if you look at the example, John Smith is our first database, and here is the result of this database. All right?
13:01
Some history of a family of Smiths. The program right now can do automatic scaling. You can say I have A4 paper or I have letter paper. Please scale the pedigree for me. It can do rotating. A lot of pedigrees, especially if you know a lot about width, but not about depth.
13:23
If you know a lot about your siblings, about nieces, nephews and so on, but you are not sure about your grand-grandfathers, so your pedigree would be wide but not too tall. It means that in many cases, programs would rotate your box to fit on a paper.
13:43
It can do rotating. It can do what I call the placing of shrubs. Shrub is a very informal term. What I want to say from the point of view of mathematicians, tree, okay, this thing is a tree, but many people say the tree is something which has a root, just one root.
14:01
If you look at this pedigree, it has two roots. It have this family, Smith family, and Brown family. And it's natural to tell it a shrub, something like couple of tree, something which have several roots but still looks like a tree. And we invented an algorithm for automatic drawing shrubs.
14:23
It was really a fun part of all this. Now, some new things. We now have automatic treating of twins, something which is not in the version which is described in graphics companion. If you see that here Jack and Mike are twins,
14:42
George, John and Jane are also twins, we can work with them. Now, a couple of words about heart problems. I would call them semi-solved. I will explain you why. The first heart problem is constant veneity,
15:02
constant veneic marriage. Constant veneic marriage are marriage between relatives, like marriage between niece and uncle here, or between cousins here. Why is it hard?
15:20
Most algorithms, and including other algorithms, maybe all algorithms for automatic placing things on a paper are recursive algorithms. You go from level to level. And if you're a mathematician, you immediately recognize it. This means that you cannot have loops because if you have loops, it's very difficult to make a good recursion.
15:42
The point of constant veneic marriages is that you have a loop. You can go to the same point but several different routes. What is our situation with constant veneity? Well, we can do it using a very dirty hack,
16:01
something really not good to talk about. What we do, we make it twice and then we delete extra nodes. So if you look at lower here, you see that it should be probably at this place, right? The reason why it's there is that the program
16:22
first make two lowers here and here and then deleted one. Again, you can manually do this. You can just take a tag file and just move here a little bit. It will work. But I just didn't want to do it here because I wanted to show you.
16:40
And this is a very simple situation. When you have many children from constant veneic marriages, you would see something really skewed. It's a bad problem but it's semi-solved in the sense that in most cases it does it right unless you are purist and you say that Jack and Laura
17:00
really should be here and here. Another problem which unfortunately absolutely unsolvable is what they call a buddhist problem. Now, unfortunately, we don't have any Polish people in the audience. We have some Lithuanian people. You might know this famous poem by Adam Mitzkevich
17:24
about old Lithuanian buddhist who had three sons, three mighty sons. He sent them, he sent one of his sons to Russia to bring rich force
17:40
and gold. He sent another son to Germany to bring amber and silver. And the third son was sent to Poland to bring the best thing Poland has, which is a Polish bride, a great Polish woman. And what happened, you can guess that they both returned
18:00
with Polish brides. It's very nice story and why it's a problem for us? Well, let's try to make a pedigree. I can put old Budrys, no problem. This old Lithuanian is here. He's his first son and the Polish wife, second son and the Polish wife, that's fine.
18:22
And then somebody will say, okay, I want to show the pedigree of all Polish brides here. I would say, okay, I will put father and mother of this wife here. It will be more difficult here because usually you want to put male to the left and female
18:41
to the right, but I can actually switch. I can take this female and put it here and put here all her father, mother, and so on. But what can we do with this woman? There is absolutely no place to put her father, mother, or anybody because it's over there.
19:03
It's a problem for manual situation. Manually people just put here the degree here and just put some lines like this. You can do this programmatically. Then you will have a lot of self intersections, a lot of problem. And I really was not able to make this work nice.
19:25
Whatever I tried to do, it looked bad. But I looked at the medical journals and then I understood that they also cannot solve it manually because all the complex pay degrees with things like this still have self intersections and they are really, really bad.
19:43
So this is a problem which is because we want to put exactly multi-dimensional degrees into two-dimensional piece of paper. And this is difficult, really difficult.
20:01
So again, this is a hard problem. Now, but still we have a progress because in the first version you could not solve it at all and in the first version you could not solve constant minute marriages either. Now we can somehow. Now what I want to do, I want to make a small demonstration
20:22
and for this demonstration I need a volunteer who wants to do. And if nobody wants, I will take either Anita or Peter because they cheer this session. Okay, what I want from volunteer, you just tell me a little bit about your family
20:42
and we will make a pay degree for you. Who wants? No, please, not. It's not that difficult and it's easy
21:01
and believe me that I will, yeah, and by the way, I will not ask you about your diseases so it will be, by ethical reasons. So it will be just something like a genealogical version of a medical pay degree, a pay degree of absolutely healthy and nice person and healthy and great family.
21:21
So, okay, come in. Okay, come here. Okay, okay, we'll start from you.
21:42
I will, I need to put any ID, let it be MA and this will be Marta. I can put the last name but then the pay degree would be too complicated so let's do. Not have it compliant. Now, now about date of birth.
22:02
It's used only to order people. If you have siblings, you go from the oldest to the youngest so if you don't want to tell me your date of birth, it's, you can tell me any year. Just make it consistent with your siblings.
22:22
Just a year? Just a year. Okay. 57. 57. 57, okay. Now, DOD is something I don't want to put here and Pro Band, Pro Band is a fancy medical name
22:45
for interviewee, for somebody who's convinced that. So you are a Pro Band, that's what I wrote. Now, let's talk about, first about your father and mother, okay? Okay. I just need name and date of birth
23:01
and whether they are with us now, so. Okay, Father Joseph. Okay, let's put Joe as his, Joseph, sorry. And he's obviously male.
23:24
I put the poem so the same read but obviously you don't want and I need his date of birth. 1916. 1916. Is he alive? No. Do you remember his date of death? 1992.
23:41
Okay, okay. And let me put here as your father, it will be Joseph. Okay, and let now go to your mother. Her name is Audrey, A-U-D-R-E-Y.
24:02
Yes, and female, male. 1915. 1915. And she is still living. Right, and let me put your mother as Audrey, right? Now, let's go to your kids.
24:23
My children? Yes, and to make it simple, let's talk only about your biological kids. If you have any adopted kids, let's not put it in the pedigree. There are some special symbols for this, but let's make it simple, okay? Daughter, Anna.
24:43
Just look that I don't have duplicating credit. Anna, and she is female. 1977. 77. And your mother is you, which is M-A, right?
25:07
And next. Son, Ross, R-O-S-S. Ross, who is male. 1979. 79.
25:23
And again, M-A, right? And son, Andy. Okay, you have already, and it will be A-D, Andy. We could put column numbers, but I usually forget which number is his male.
25:44
1982. 1982, the same. Okay, now let's talk about your siblings. Oldest is Ruth, 1940.
26:11
1940. Living. Mm-hmm, and it's Audrey and Joseph, right? Yes.
26:20
You can have half siblings, which share only father or mother with you. Two siblings, so next is Bernard. Bernard, that's me, my old. 1942.
26:40
1942. He is deceased. Okay, do you know the date of his death? 1964. 1964. 64, let's do it this way. Okay. Next is Mary. Okay.
27:08
1944. 1944. And she's deceased in 19, oh, 2006.
27:22
Okay. Next is Gerald. Gerald. 1948.
27:41
Mm-hmm. Right? And next is Mark. You have a really good family. German family. Yes. Male. 1958. 1958, and he's living.
28:01
Yes. Okay. This one? Yes. Now, Leila, what should we go next to the... Children of your siblings. Children, so your... Beginning with the oldest. Yes. Okay. Yeah, we can stop at any moment. If you... They're not too many.
28:21
Okay. Okay. Martha. Okay. Which symbol we still have? Empty. Martha. That's a child of Ruth. Male. Do you remember who that? 1968. 68.
28:41
And first is mother, and Ruth is our youth for us, right? Please check what I'm typing. I'm famous for my typos. Okay. Next is Sarah. Sarah, female.
29:01
1975. 75. And she's... Child of Mary. What is... Mary M.R., right? And next is Helen. Mm-hmm. Female. 1976.
29:20
76. She's daughter of? Mary. Mary, also Mary, right? Also Mary. Yes. Okay. Next is Leonard.
29:45
Mary. 1977. Mm-hmm. And he's a kid of? Gerald. Okay, Gerald is G.E., right? Yes. Okay. Next is Joel.
30:05
Joel. Joel. J-O-E-L. Yes. J-O-E-L. Mm-hmm. Male. Oh, I cannot do... Thank you very much, because otherwise you would have a really big problem.
30:20
1982. 82. And he is a kid of? Gerald. And Gerald? G.E. G.E., okay. Oh, sorry, sorry, okay. It's, yeah, you're absolutely right.
30:46
You're absolutely right. Thank you very much. Like this. Is it the only problem I have here or something else? No? Okay. One more. Curtis.
31:00
C-U-R-T-I-S. Curtis. And he is male? 1984. 1984. And he is kid of? Gerald. Also father, right? G.E., okay. That's it. That's it. We can stop here. We can go more and more,
31:22
and if you ever go to a doctor, it would be really much more, but let's see what do we have here. Okay. Yes. First we'll make a tag file. Okay. Here's a tag file we generated.
31:44
And let's try to make a PDF file. PDF. Again, I'm sorry, it's Emacs and command line. X-P-T-L-A-X-P-D-F.
32:05
Let's see what do we have here. Okay, it wants us to rotate it. Because, let's see. Okay. Let me increase it a little bit.
32:27
Probably 150 is the best. Okay, you can take a look and tell me whether you made a mistake. You have a marker and you have an arrow, and the arrow means that you are a pro band. We started with you. You have your father and mother,
32:44
and this looks like a pedigree, unless we made a mistake here. Did we? And it's not very much complicated because we don't have any marriage between your cousins and so on, so it was very good in making it simple.
33:00
And if you go next page, it's actually, let me rotate it back. Rotate it counterclockwise. It's actually, it has a legend. It tells you who was there, when they were born, and whatever was in our pedigree.
33:21
As you see, it's very simple, and I would bet that if you would try to do it manually, you would spend much more time here. Okay, thank you very much, and if you wish, if you have memory stick, I can give you this pedigree as a gift. Okay, questions?
33:40
Comments? Yes? You see me? Okay, it can work because it's tech. It can work with anything you want, and here, actually, we have two,
34:02
right now, we have two models, English and Russian. I can show you. I don't know, but they have it anywhere compiled here. Let me see. Let me see. No, this is also English.
34:22
We have here, okay, CSV. Yeah.
34:42
Okay, and now you can see, yes. Here is, yeah, it's very garbled because it's, I did some strange thing with this,
35:01
but if you take a look, all names are in Russian, and it's obviously extensible, so if you can, if you want to extend it to Czech or Arabic or Hebrew, no problem with this. Just put the proper Babel reference.
35:21
Yes? Is there any other way, are these the only ways? Are they fully manual or using this system? Is there any other automated or semi-automated way of creating these diagrams now? Well, there are several programs which do this, and most of them are, I actually forgot to tell you about the last slide here.
35:43
What you want to do, most of programs are interactive. You just have a mouse, you put a point here, click here, click here. There are several commercial programs now available, and because the main audience for them are doctors,
36:00
and doctors are supposed to be rich, these programs usually cost a lot of money. What we want to do, and we probably will do, we will do it semi-mouse-driven. So it will make a pedigree like this, and then you can take a mouse and shift and adjust, and it will make it, that's what we are going to do.
36:20
But if you have a lot of money, there are some programs I can recommend to you which, I think they are less convenient than what we are doing, but I'm biased obviously, but our program is free. I have a lot of money, but I don't have any interest in this. Yes?
36:44
Excuse me? Mr. Polish Wives. With what Polish Wives? Mr. Polish Wives. Ah, the Polish Wives picture, yes. Surely, here. No, I know that this one is good.
37:01
Perhaps this one is good too, because if you have a wife that is married to her father, 80, and then divorce, and then marriage. Ah. And you get, immediately, another look, how do you express this? It's, you mean, what happens if a wife has first marriage to the first brother, and then to the second?
37:20
Then usually you take two lines, which shows two marriages. Yeah, and then you just show the kids of the first marriage, the kids of the second marriage. There is nothing conceptually more complex than the standard constant meaning marriage is problem. You just have a look. So, it works with this.
37:41
And you can divorce. Yeah, and you can put divorce by breaking the line. It knows this. Yes? In bookshops, we usually use labels, in books which are big diagrammed in different parts. Yes. Can you do that in this standard? Manually, yes. But the problem, yes. One of the way of solving this problem of complex pedigree is to make parts,
38:03
and then make C page next, or C part two, and C page three. And then you have the same problem as we have this breaking displayed equations into lines. Human can do this. Programs are really bad in logical breaking
38:21
of big chunk of information into logically consistent parts. Yes, manually it's easy to do. I don't know how to do it absolutely automatically. Okay, let me, ah, yes.
38:45
Excuse me? You are trying to visualize the probability for the evidence of a process. Mm-hmm. I would think that the, some of the behavior is some kind of graph,
39:00
and what I'm interested in is can you feel this graph for a specific process? Yeah, but the graphs here are not trees. The pedigree is not a tree. It's a general graph, which is the only thing it has, the only thing it usually has something, it has generations.
39:20
You want it to be layered as generation. But the biggest problem, okay, if it were a tree, I wouldn't have 90% of my problems because it's not a tree. It's a general, it's a graph of a general, it's a general graph. You are absolutely right, and that is why it's so hard to do.
39:47
Get it? What? Yes. You can put the graph on paper. Yes. And not to put it in text. Yes, yes, yes, yes, yes. Fortunately, this is a little bit less general graph,
40:02
so we can do some solutions for this version. But the thing is, the general problem getting the graph on paper is, as everybody knows, unsolvable. You cannot do it in a general situation. Yes, yes, yes. And this problem here is actually part of this.
40:21
In general case, you cannot solve it. That's how God created this world. Any other questions? Yes? In the term of the generation of the children.
40:42
I go, the generation of a kid is one plus a generation of the youngest parent. And I think it's the only rational way of doing this. Any other questions? Okay, then I want to, at the end, I want to thank,
41:04
okay, my work was three in this project, but my collaborator actually got very nice travel grants and lots of grants from Russian sources, and I want to thank all of them. And I want to thank Tech Users Group, which you helped visit with the grants for this.
41:24
And I want to help all you for being a nice audience. And I promise that in a year or something, in the next conference, I will show you some even more nice tricks to doing with a degree. We can do it much more interactive, much more,
41:43
you see me? User friendly. Much more user friendly, and we want to make it really good. Thank you very much.
Recommendations
Series of 21 media