WEBVTT 1 00:00:00.840 --> 00:00:01.050 Julie Hogan: you're. 2 00:00:02.550 --> 00:00:16.650 Julie Hogan: able to hear, so what we're going to do is walk through two other two other objects that are slightly more complex in their treatment than electrons and nuance and that's jetson missing transverse energy. 3 00:00:17.580 --> 00:00:36.120 Julie Hogan: So the main goal is to kind of do the same type of walkthrough of how jetson met are treated in cms specifically in poet, and then we're going to actually do our exercises on an on an element of JET analysis that became much less a much simpler, more. 4 00:00:38.250 --> 00:00:44.820 Julie Hogan: More user friendly in 2015 data which is now available to you on the open data portal and that's jet substructure. 5 00:00:45.540 --> 00:01:01.200 Julie Hogan: So let's dive into the to the beginning of the lesson here, this is the jetson met episode of the physics object lesson and so, hopefully, many of you are already able to get this blue box working if you are on zoom especially. 6 00:01:02.160 --> 00:01:10.980 Julie Hogan: What we're going to do is use the main branch of poets now as opposed to the exercise branch that you were working with before. 7 00:01:11.910 --> 00:01:32.760 Julie Hogan: So the instruction is to go to this to the same area you've been using get stash your changes and check out the mini O D branch, so that you have the the let's say the full poet sweet as opposed to the mini workshop branch. 8 00:01:34.020 --> 00:01:42.990 Julie Hogan: And so what I would like you run, this will probably take the entire you know, most of the time that i'm speaking to you what I would like you to run is a test file, that is a. 9 00:01:43.920 --> 00:01:58.350 Julie Hogan: Top court pair simulation with high mass so it's talk work pairs that have unusually high mass and so, in order to do that we need to make two edits in the configuration file that you studied before with. 10 00:01:59.520 --> 00:02:13.710 Julie Hogan: So the first the first edit we're going to make is that we're going to actually process all the events, so we can do that by saying negative one to run over all the events in this particular line called Max events. 11 00:02:14.790 --> 00:02:33.030 Julie Hogan: And i'm also going to ask you to change the input simulation file so from the the workshop tutorial site here, you can come and copy this actually we could I suppose copy the entire thing you can copy this little element here, and then in your favorite. 12 00:02:34.350 --> 00:02:39.870 Julie Hogan: text editor mine is emacs design crazy according my students. 13 00:02:41.970 --> 00:02:46.530 Julie Hogan: You can come and put in those pieces of information. 14 00:02:47.910 --> 00:02:52.740 Julie Hogan: So this is the big poet config. 15 00:02:53.790 --> 00:03:07.920 Julie Hogan: And so we're going to change in in two places one of them is right here process dot Max events, you can see of course i've already done it because I already ran this file before in preparation. 16 00:03:08.550 --> 00:03:28.500 Julie Hogan: But you can make that same negative one, and you could come in here, and you could paste in what you copied from the web, so that you have this high mass TT bar route file ready to process, which is a TT underscore mass of the Tiki bar pair 1000 to infinity. 17 00:03:30.480 --> 00:03:33.420 Julie Hogan: i'm going to undo that because I didn't actually need it. 18 00:03:36.930 --> 00:03:37.230 Julie Hogan: Okay. 19 00:03:38.310 --> 00:03:46.770 Julie Hogan: So hopefully your your config looks something like this and you have this root file ready to run over all of the events. 20 00:03:51.960 --> 00:03:53.520 Julie Hogan: And then you can run that. 21 00:03:58.320 --> 00:03:59.640 Julie Hogan: By saying. 22 00:04:01.680 --> 00:04:02.130 Julie Hogan: whoops. 23 00:04:03.180 --> 00:04:16.530 Julie Hogan: We don't even need to say true is the default is money Carla now Edgar help me out real quick Do I need to instruct people to scram or with their existing compile of the fat jet analyzer be okay. 24 00:04:28.800 --> 00:04:33.390 Julie Hogan: Are you all still able to hear me I don't know if you're was able to hear that question I didn't hear any response. 25 00:04:38.640 --> 00:04:43.440 Matthew Bellis (he/him): I think he just didn't have his microphone on I think I think he was muted, but I think he heard you. 26 00:04:45.510 --> 00:04:46.740 Julie Hogan: So do you think I should. 27 00:04:46.830 --> 00:04:47.940 Julie Hogan: Talk people to scram. 28 00:04:48.570 --> 00:04:50.250 Edgar Carrera: yeah I think you should you should. 29 00:04:50.640 --> 00:04:53.460 Edgar Carrera: Do this crime shows they they checked out. 30 00:04:54.240 --> 00:04:59.700 Edgar Carrera: The new version, the new a new branch, so it should be kumbaya. 31 00:05:00.600 --> 00:05:08.100 Julie Hogan: Okay, so let's add an instruction in there once you've done the get checkout please scram my scramble look really. 32 00:05:08.100 --> 00:05:08.610 Edgar Carrera: silly. 33 00:05:09.270 --> 00:05:10.860 Julie Hogan: it's nothing as needs to be done. 34 00:05:12.540 --> 00:05:15.360 Julie Hogan: But please scram and then you can run. 35 00:05:16.920 --> 00:05:17.790 Julie Hogan: run poets. 36 00:05:19.590 --> 00:05:28.860 Julie Hogan: Just like that Okay, since we lost a little bit of time i'm going to carry on, and so the blue box is hopefully we'll just run in the background, while we talk about what's going on with jets. 37 00:05:30.390 --> 00:05:36.840 Julie Hogan: Okay, so, as I was mentioning the jets in the missing transverse energy are a little bit more. 38 00:05:36.960 --> 00:05:39.240 Julie Hogan: Complex objects, because they they. 39 00:05:41.040 --> 00:05:51.390 Julie Hogan: They depend on many, many, many, many particles so jets as as many of you will know if you have a high energy physics experience their their groups. 40 00:05:52.500 --> 00:06:05.220 Julie Hogan: hydrants groups of how drones or other or photons or other similar particles that are produced, when a Cork or blue on emerges from a collision and hadron. 41 00:06:06.630 --> 00:06:26.100 Julie Hogan: And so what we observe and cms is a shower of particles that have sort of a Cone like shape and the energy of all the particles contained in that shovel should informs us about the energy and, of course, the direction and everything of the initial particle that forms the JET. 42 00:06:27.270 --> 00:06:34.440 Julie Hogan: And so there can be hundreds of particles in the actual shower they tend to be a mixture of charge hadron. 43 00:06:34.860 --> 00:06:47.040 Julie Hogan: you'll have a lot of photons because neutral pions will be produced in in hadron decays and then they'll produce photons and we can get neutral hadron as well, and so these are very, very messy objects. 44 00:06:48.210 --> 00:07:02.790 Julie Hogan: And we have to figure out how to take all of the energy deposits in the detector and and create jets out of them it's not really a one to one thing like like certainly like new ones are. 45 00:07:04.260 --> 00:07:18.960 Julie Hogan: So we use clustering algorithms to produce to produce jets from energy deposits in the detector There are a variety of ways to do this, many experiments just use calorie emitter energy deposits. 46 00:07:19.860 --> 00:07:26.610 Julie Hogan: You can certainly cluster simulated information, you can use generated particles and do clustering. 47 00:07:27.030 --> 00:07:36.540 Julie Hogan: But in cms what we use for analysis, are the particle flow candidates and Edgar introduced the idea of particle flow at the beginning, this morning. 48 00:07:36.900 --> 00:07:56.970 Julie Hogan: And so particle flow essentially says here are all of the neutral hadron the charge hadron the photons the electrons nuance here all of those candidates that are reconstructed in the detector and then from those particle candidates, we can perform clustering to find jets. 49 00:07:58.650 --> 00:08:10.170 Julie Hogan: And so we do that by taking initial particles so let's say this is a representation of particle flow candidates with the larger bubbles being larger energy. 50 00:08:10.890 --> 00:08:25.020 Julie Hogan: And we start a process of comparing two particles together and finding their distance and then we have a criteria based on based on energy relationship or momentum relationship. 51 00:08:25.560 --> 00:08:49.740 Julie Hogan: And that distance, so we have this criteria for combining particles together and we proceed clustering things that past this condition we proceed proceed proceed until more and more of the particles have been combined together if they share if they're in close proximity. 52 00:08:50.970 --> 00:09:12.030 Julie Hogan: and eventually we have a description of all of the energy deposits in the event, based on a certain number of jets and then within a jet we have many, many constituents, so in this little example you can you can map oh my screen is just tiny bit too big. 53 00:09:13.200 --> 00:09:29.640 Julie Hogan: There you can map all of this initial energy up here in the top left with the JET constituents in the bottom right, and so we would say that this event has four jets and each jet has a number of constituents. 54 00:09:31.170 --> 00:09:47.160 Julie Hogan: And so, this clustering is done done automatically when jets are reconstructed and we use the fast jet package and we use the anti katie algorithm in order to choose what the rules are for combining energy deposits. 55 00:09:48.450 --> 00:09:55.830 Julie Hogan: And so the anti katie algorithm if you kind of visually map things out we get jets that are essentially circular. 56 00:09:56.490 --> 00:10:06.810 Julie Hogan: they're essentially circular and the high energy deposits tend to be relatively well centered along what we would call the axis of the JET. 57 00:10:07.650 --> 00:10:19.590 Julie Hogan: And so that's kind of what this little graphic represents for you can see, the higher energy deposits within the usually usually pretty well centered here on these circular jet clusters. 58 00:10:21.630 --> 00:10:40.170 Julie Hogan: So this is fine, this is great, except for the fact that we have a lot of energy deposits in our in our detector that are not coming from the actual proton collision they come from other proton collisions or whatever else, and we call that pilot. 59 00:10:41.190 --> 00:10:55.500 Julie Hogan: So we have to have algorithms to deal with pilot and and, again, there are many ways to do this, but the two that are the most significant for cms are called charge hadron subtraction which we abbreviate to see hs. 60 00:10:56.130 --> 00:11:03.060 Julie Hogan: And then the more whimsical pilot per particle identification algorithm which is abbreviated puppy. 61 00:11:04.380 --> 00:11:10.710 Julie Hogan: And so we're going to focus on charge hadron subtraction but in 2015. 62 00:11:11.430 --> 00:11:27.840 Julie Hogan: We don't have this information available in poet, yet, but in 2015 you can access the puppy information so we'll try to build that out later, or if you're really interested in it feel free to ask us that might be motivation to include it, I became kind of the default in 2016. 63 00:11:29.520 --> 00:11:42.840 Julie Hogan: So charge hadron subtraction where it's like this charge had ron's by definition in the particle flow algorithm have a track associated with halloran that are energy deposits and so that track. 64 00:11:43.800 --> 00:12:05.490 Julie Hogan: identifies a vertex and so all the tracks that are not part of the that are not emerging from the primary vertex can be automatically labeled as coming from pileup and so energy deposits that are linked directly to those pileup tracks can be removed from the from the JET as pilot. 65 00:12:06.510 --> 00:12:09.810 Julie Hogan: And so that's one of the one of the automatically. 66 00:12:10.830 --> 00:12:13.260 Julie Hogan: Applied algorithms for cms jets. 67 00:12:15.180 --> 00:12:32.940 Julie Hogan: Okay So how do we access them that's the little bit of background on jets, how do we access them if you have participate if you happen to participate in this workshop before you'll know that when we were talking about 2011 and 2012 open data, we had to we had several levels. 68 00:12:34.260 --> 00:12:40.650 Julie Hogan: We had different ways, you could access the jets than a different pros and cons, this is much simpler now with the. 69 00:12:42.030 --> 00:12:47.580 Julie Hogan: Open data which is all in this mini O D format, because everything uses this pat. 70 00:12:48.720 --> 00:13:00.420 Julie Hogan: Reconstruction that Edgar introduced you choose physics analysis toolkit infrastructure and so now in poet, we just have one jet analyzer in the branch. 71 00:13:01.980 --> 00:13:07.800 Julie Hogan: Where we can work with sort of normal jets and then we'll look in the next session at our that jets. 72 00:13:09.450 --> 00:13:13.260 Julie Hogan: So, if you look at the JET analyzer. 73 00:13:14.820 --> 00:13:16.830 Julie Hogan: You will see all the basic stuff. 74 00:13:19.890 --> 00:13:22.950 Julie Hogan: It of course lives in the source folder there we go. 75 00:13:24.930 --> 00:13:37.320 Julie Hogan: So you'll see all the basic stuff branch branch decorations like you saw in your other analyzers and then, when we get to the main function, which is called the analyze. 76 00:13:40.830 --> 00:13:55.170 Julie Hogan: We open up a collection of jets so i'm i'm online 343 when I find this thing that i've linked in the in the tutorial where we open up our list of pat jets and we begin to loop over them. 77 00:13:57.990 --> 00:14:02.790 Julie Hogan: And we do a whole bunch of processing, but if you if you scroll down. 78 00:14:04.200 --> 00:14:26.010 Julie Hogan: To more like line 404 you'll see that we can use the same type of functions energy PT Ada by charge mass the xp y Z you can use all the same sort of basic variable access methods on a pat jet that you saw for electrons and. 79 00:14:27.330 --> 00:14:32.040 Julie Hogan: Those are completely common completely common to cms objects. 80 00:14:33.540 --> 00:14:38.610 Julie Hogan: So we can access all of the basics of the JET using those methods that you're you've already seen. 81 00:14:40.050 --> 00:14:54.480 Julie Hogan: um but for jets, we need to, we need to start being a little bit more clever and a little bit more prescriptive when we talk about things like identification variables so for the electrons for the nuance these were just variables that are calculated and stored in poet. 82 00:14:56.040 --> 00:15:15.780 Julie Hogan: here for jets we've gone ahead and done it for you, where we've applied what we call the noise jet ID jets can emerge just because there has been a noise event in the calorie meter, and so we have developed an algorithm to get rid of all that for you. 83 00:15:17.130 --> 00:15:23.160 Julie Hogan: And we apply that automatically in the config so i'm going to go back to the poet config. 84 00:15:24.930 --> 00:15:26.040 Julie Hogan: and show you this. 85 00:15:27.870 --> 00:15:40.170 Julie Hogan: The top part here looks pretty familiar like what you saw before, and then we get down to the poet analyzers and you have the electron in the new on analyzer the TAO the photons. 86 00:15:41.190 --> 00:15:55.470 Julie Hogan: And then we start to kind of get into it projects there's a very there's a lot of configuration going on to deal properly the jets and the first segment of it is this one apply the noise jet ID filter. 87 00:15:57.450 --> 00:16:02.670 Julie Hogan: And so cms is provided this function for you, where, when we run poets. 88 00:16:03.990 --> 00:16:14.610 Julie Hogan: These lines, right here we'll we'll call call the noise jet ID filter for both our small radius jets and our large radius juts. 89 00:16:16.050 --> 00:16:22.440 Julie Hogan: So after showed you how to dig a little bit in the cms code, if you are really interested in seeing all of this specific. 90 00:16:23.190 --> 00:16:34.710 Julie Hogan: criteria that go into the noise ID you can go and search for this guy the PF jet ID selection filter and you'll find all of the criteria that it's applying for the jets. 91 00:16:41.280 --> 00:16:51.420 Julie Hogan: So those are the super basics of working of working with jets the super basics missing transverse energy is very closely related. 92 00:16:52.740 --> 00:17:09.330 Julie Hogan: It does go beyond Jesse it's a way of looking at all of the particle flow candidates in an event so missing trends or so momentum is the negative vector some the transverse momentum of all the particle flow candidates basic definition and so. 93 00:17:10.530 --> 00:17:17.340 Julie Hogan: We call the the magnitude of that we call the met missing transverse energy okay not not quite. 94 00:17:18.510 --> 00:17:24.870 Julie Hogan: Great grammar syntax for that, but that's what we call it and we abbreviate it Max and that's what you'll see all over cms code. 95 00:17:26.010 --> 00:17:26.910 Julie Hogan: So. 96 00:17:28.620 --> 00:17:39.030 Julie Hogan: This is negative vector some is of course very easy to compute and it's usually in its raw form not entirely helpful because we don't. 97 00:17:39.510 --> 00:18:00.960 Julie Hogan: measure the momentum of all of these particle flow candidates with with super correctness, we have to apply corrections to basically to basically everything, and particularly the jets so we have a method, called the not not very creatively called the type one correction. 98 00:18:02.940 --> 00:18:20.400 Julie Hogan: For missing transverse energy and the goal of this correction is to say okay I started with all the particles locanda certain my jets but then I have to make energy corrections for the jets to get them right, so let me propagate that energy correction to this missing transverse energy. 99 00:18:21.540 --> 00:18:23.580 Julie Hogan: And so that excuse me. 100 00:18:24.960 --> 00:18:31.200 Julie Hogan: That is also done for you and you don't have to worry about it anymore, other than to know that it has been done for you. 101 00:18:32.700 --> 00:18:34.890 Julie Hogan: So in the config i'm going to skip down. 102 00:18:36.930 --> 00:18:41.400 Julie Hogan: i'm going to skip down to the missing each section right here. 103 00:18:42.630 --> 00:18:46.200 Julie Hogan: propagate the JET energy corrections to the missing et. 104 00:18:48.030 --> 00:18:59.790 Julie Hogan: And we do that by first explicitly evaluating the raw or completely uncorrected missing transverse momentum that's what these lines are responsible for. 105 00:19:01.080 --> 00:19:03.330 Julie Hogan: And then we evaluate the correction. 106 00:19:04.500 --> 00:19:09.390 Julie Hogan: So we have to basically prepare this correction that that's this. 107 00:19:11.310 --> 00:19:13.950 Julie Hogan: method here called Type one core for new jack. 108 00:19:15.420 --> 00:19:30.690 Julie Hogan: We have to evaluate a correction and then apply it, so an ED producer, in contrast to immediate analyzer and ED producer is creating some information to store it into the into the events they are processing. 109 00:19:33.510 --> 00:19:44.310 Julie Hogan: So all that has been done for you and in the poet met analyzer you get to just open up this newly created met package and go on with it. 110 00:19:46.410 --> 00:19:46.770 Julie Hogan: alright. 111 00:19:48.120 --> 00:19:57.600 Julie Hogan: So, the last thing i'm going to say here about as a basic intro to the mat is to look at to look at its analyzer. 112 00:20:03.570 --> 00:20:13.950 Julie Hogan: Okay, this analyzer is much shorter and looks a bit different from the from the electronic new on analyzers that you've already seen. 113 00:20:15.030 --> 00:20:24.930 Julie Hogan: Primarily because the missing transverse momentum is is one value per event so there's Actually, this is it the analyze function fits in you know 20 lines. 114 00:20:26.190 --> 00:20:31.140 Julie Hogan: And so there's nothing to loop over, we just need to ask for the pat met. 115 00:20:33.600 --> 00:20:40.800 Julie Hogan: You to ask for for the path net and then we can access several interesting variables, we can actually get the some. 116 00:20:41.190 --> 00:20:49.770 Julie Hogan: Some etc, so this is the the total transfers energy recorded in the event, so it gives you a basic measure of how much energy was was recorded. 117 00:20:50.400 --> 00:20:56.220 Julie Hogan: We have the actual missing transverse energy magnitude net.pt. 118 00:20:57.030 --> 00:21:05.940 Julie Hogan: And then we can get the X and y components as well, and the fly, we obviously cannot get the Z components, because we just don't have that information we don't have. 119 00:21:06.420 --> 00:21:15.600 Julie Hogan: We don't have enough instrumentation in the very, very forward regions, in fact, we would need it, even within the beam pipe basically in order to get an accurate reading. 120 00:21:16.320 --> 00:21:24.930 Julie Hogan: For the Z direction of this variable so we're only working in the transverse plane when we add up how, when we add up this missing energy. 121 00:21:27.060 --> 00:21:34.650 Julie Hogan: And you can also get what's called the significance, so the significance of the missing transverse energy is very similar in its usefulness. 122 00:21:35.460 --> 00:21:40.260 Julie Hogan: To the tracking to the track significance, you are extracting earlier. 123 00:21:40.740 --> 00:21:54.780 Julie Hogan: For electrons it basically gives you a way to judge whether the missing transverse energy and this event is probably real from invisible particles that were not detected like neutrinos or dark matter or susie or whatever. 124 00:21:56.400 --> 00:22:13.740 Julie Hogan: or versus just just instrumentation error, where you know, this is just a, this is just a Z bows on decaying to leptons and so any missing energy is telling you there's some this measurement and so you can judge between those things, using that significance. 125 00:22:15.600 --> 00:22:31.920 Julie Hogan: And then just for your information if you're ever using this we've also included for you the raw values, so the raw energy and PT and five values, so you can get a handle on how how much these values changed when widens corrections. 126 00:22:34.260 --> 00:22:34.650 Julie Hogan: Okay. 127 00:22:36.030 --> 00:22:40.350 Julie Hogan: that's a quick look at the at these basic objects. 128 00:22:41.640 --> 00:22:55.380 Julie Hogan: So I suppose we can take you know, like one minute or so for questions if anyone has them or if anyone failed to to do the blue section of the top before we go to the interesting that. 129 00:23:07.560 --> 00:23:10.530 Julie Hogan: I don't hear anything so i'm going to click the button. 130 00:23:11.760 --> 00:23:12.420 Julie Hogan: And here we go. 131 00:23:14.040 --> 00:23:21.000 Julie Hogan: um so let me show you are trying to hit a different button i'm going to go back to the menu i'm just going to show you what exists in this. 132 00:23:21.930 --> 00:23:29.310 Julie Hogan: lesson we're going to look here at substructure, we are not going to look at the tagging. 133 00:23:29.820 --> 00:23:36.480 Julie Hogan: we're not actually going to look explicitly together synchronously at the JET corrections, we were just talking about. 134 00:23:37.140 --> 00:23:49.830 Julie Hogan: For the Miss when we when we talked about the missing transverse energy but they're documented here, so you can read them on your own we're going to kind of skip those things and we're going to look at the joke substructure because it's new. 135 00:23:51.060 --> 00:23:54.780 Julie Hogan: So this blue box is the same as the previous blue box i'm gonna skip it. 136 00:23:57.660 --> 00:24:10.380 Julie Hogan: And one of the things that we spend well I feel like I spend most of my time thinking about this because I searched for new physics that basically always produces really interesting jets. 137 00:24:10.830 --> 00:24:21.750 Julie Hogan: And so, when when we're looking at a jet you there's a huge variety of things that could come from it, it could be a light court usual regular quark or glue on. 138 00:24:22.200 --> 00:24:32.610 Julie Hogan: Even those things look different from each other, it could be a heavy flavor bark it could be a completely fake jet that's from pile up, it could actually even. 139 00:24:33.120 --> 00:24:47.730 Julie Hogan: The JET could contain an entire heavy standard model particle and all of those different sources, create a jet that looks different, even though we keep them all in the same collection and call them jets. 140 00:24:48.840 --> 00:25:04.560 Julie Hogan: And so, one of the one of the really interesting things we can do is try to use algorithms to peace apart what's inside the JET and determine what I came from so we're going to talk briefly about jet mass and jet substructure. 141 00:25:06.030 --> 00:25:15.150 Julie Hogan: So the massive a jets, this is a picture of an example jet so so everything within the orange circle might be energy deposits that have been clustered in a one jet. 142 00:25:16.350 --> 00:25:23.970 Julie Hogan: And you can see right away from the three purple circle so we've drawn on that those energy deposits clump themselves in three zones. 143 00:25:25.800 --> 00:25:36.150 Julie Hogan: And the massive this jet would would come from basically Taking all these four vectors of all of the particle flow candidates that are within the orange circle and adding them together. 144 00:25:38.040 --> 00:25:49.710 Julie Hogan: And that brings with it, some distortion, so it in a the extreme example I could I could observe a chat from a glue on me and we know who wants to the masses. 145 00:25:50.340 --> 00:26:05.070 Julie Hogan: But if I measure the mass of that jet i'm not going to get zero i'm going to get some some hopefully small number, but it will definitely be non zero because we have all of this radiation out from our away from our. 146 00:26:07.080 --> 00:26:25.950 Julie Hogan: kind of central axis of our jet and all of this radiation basically falsely increases the observed mass of this object, and so we see this when we try to study jets from things where we know the mass like w bows ons or top marks which spoiler alert, this is a tough work. 147 00:26:28.650 --> 00:26:31.860 Julie Hogan: We observe that there's there's a lot of kind of. 148 00:26:32.880 --> 00:26:44.580 Julie Hogan: background contamination let's say if we if we try to say, let me, let me find all the top works that are near 170 gv there'll be a whole bunch of light working through on jets in that. 149 00:26:44.970 --> 00:27:00.120 Julie Hogan: In that space and the top court jets will have a mass that's kind of wide and then distorted and so we can do better at representing the mass of the initial particle that created the JET by applying grooming algorithms. 150 00:27:01.350 --> 00:27:03.210 Julie Hogan: And there are a whole bunch of them. 151 00:27:04.290 --> 00:27:22.350 Julie Hogan: There are a whole bunch of them and cms uses primarily primarily nowadays we use the soft drop algorithm but in 2015 the pruning algorithm was also very popular and it's effectively, these are algorithms that have been designed. 152 00:27:23.460 --> 00:27:39.900 Julie Hogan: To go through the JET and kind of cluster it again repeat the clustering process or go backwards and D cluster and at each step study, where the radiation appears, is it very far from the JET axis is it very soft. 153 00:27:41.040 --> 00:27:51.870 Julie Hogan: And so things that are scattering off at a at a at a wide angle, or that have very low momentum fraction they're basically groomed out. 154 00:27:52.560 --> 00:28:06.660 Julie Hogan: So we take our pets to the groomer and we give them a you know we clip their leash and we give them a haircut and we trim their nails and we do all these things so that they, you know resemble the breed standard of whatever jet they're supposed to be much more closely. 155 00:28:08.280 --> 00:28:08.880 Julie Hogan: In the end. 156 00:28:10.380 --> 00:28:11.880 Julie Hogan: And so, in in the. 157 00:28:12.930 --> 00:28:17.220 Julie Hogan: mini oh gee we have some of these groomed masses stored for you. 158 00:28:17.910 --> 00:28:28.080 Julie Hogan: And what we're going to look at which was probably different compared to the electrons and the nuances we have these user float methods on our path objects. 159 00:28:28.710 --> 00:28:45.030 Julie Hogan: and user float is exactly what it sounds like is a float that has been inserted in the object by a user someone has engineered that this calculation occur and they've plugged the answer in kind of manually to the JET object and that allows us to pull it out. 160 00:28:46.290 --> 00:28:56.550 Julie Hogan: So if you open up the fat jet analyzer i'm not going to do it for the right now live for the sake of time if you open up the fascia analyzer you'll be able to. 161 00:28:56.550 --> 00:28:57.960 Julie Hogan: find some of these. 162 00:28:58.050 --> 00:29:12.420 Julie Hogan: fruit, in particular the prune mass and the soft drop mass stored for you stored for you, using user floats, and so we have those in your poetry. 163 00:29:14.610 --> 00:29:15.540 Julie Hogan: And so. 164 00:29:16.650 --> 00:29:35.400 Julie Hogan: Since we got a little bit behind I think what's probably best is, for we have some exercises and it's probably best is for me to carry on and then have everyone choose one of the exercise boxes, that they would like to do since we lost so much time because I don't want to go over. 165 00:29:36.810 --> 00:29:46.980 Julie Hogan: So i'm going to carry on and and introduce a couple more user floats which are the end readiness or in subject enos variables. 166 00:29:49.560 --> 00:29:57.090 Julie Hogan: The code lines look exactly the same here we're taking a jet object and we're using the user float. 167 00:29:58.230 --> 00:30:05.010 Julie Hogan: to access a specific string which is called angelina's aka TAO one toe to Intel three. 168 00:30:06.480 --> 00:30:09.150 Julie Hogan: And what these stand for the towel variables. 169 00:30:10.860 --> 00:30:29.040 Julie Hogan: The towel variables for the end subject enos function so i've linked the paper here in fact i'm realizing live right now that the equation, the equation that is supposed to appear right here is not appearing so let's click on the PDF and find it. 170 00:30:32.430 --> 00:30:45.240 Julie Hogan: there's a great paper that you can use and it defines here they are a set of town variables which depend on their their basically a PT weighted. 171 00:30:46.290 --> 00:31:02.220 Julie Hogan: base averaging That was a wacky way to say it, but what you're what you're looking at is minimum separation between jet elements, so all these delta ours are separations and those separations are weighted by the relative momentum and added up. 172 00:31:03.510 --> 00:31:09.180 Julie Hogan: And so you can get these towel variables, for whatever and you choose. 173 00:31:10.230 --> 00:31:10.830 Julie Hogan: and 174 00:31:12.150 --> 00:31:29.910 Julie Hogan: And is basically describing how many what's the maximum number of subjects that fit in this jet and so a subject, just so that you have a mental picture, a subject is one of these purple circles within a large within a large orange circle of a full jet. 175 00:31:32.670 --> 00:31:33.450 Julie Hogan: So with. 176 00:31:34.470 --> 00:31:34.950 Julie Hogan: With. 177 00:31:35.220 --> 00:31:47.160 Julie Hogan: groomed masses and with substructure information like angelina's we're in subject enos we can actually do a lot to learn what's inside the JET. 178 00:31:49.170 --> 00:32:07.200 Julie Hogan: The other substructure information we have available to us is information about the flavor of the subjects, so we can actually ask for the subjects themselves, this is definitely new again we're using that kind of creepy auto but I like it. 179 00:32:08.220 --> 00:32:15.180 Julie Hogan: In this case, we're getting a collection of subjects from our large radius jets. 180 00:32:16.230 --> 00:32:23.790 Julie Hogan: And they're the soft drop algorithms subjects and there tend to be two of them, usually this numerical value will be too. 181 00:32:24.480 --> 00:32:37.680 Julie Hogan: And so we can ask each of them for their be tagging algorithm discriminate which is described pretty well in the next in the extra episode that you can read offline and we can also ask for their flavor. 182 00:32:38.700 --> 00:32:42.810 Julie Hogan: which will be meaningless if it's data, but it will, it will have information if it's been incredible. 183 00:32:44.520 --> 00:32:55.680 Julie Hogan: And so we can get information about the number of subjects or information about the flavor and that will that will let us figure out what type of what type of JET we have. 184 00:32:56.910 --> 00:33:14.190 Julie Hogan: And so cms has combined these into algorithms kind of criteria to describe a w jet and a top jet and those were the those were the supported ones in 2015 Of course there are many others, and so you can read here. 185 00:33:15.300 --> 00:33:25.290 Julie Hogan: And you can read here in the call outs what the criteria are for selecting a w goes on, or for selecting a top work. 186 00:33:27.510 --> 00:33:47.220 Julie Hogan: And so that's the that's the information, and here we have three three exercises if you were able to run that high mass TT bar you could choose an exercise either to look at how the groomed mass relates to the particles momentum. 187 00:33:48.810 --> 00:33:54.300 Julie Hogan: Or how the mass relates to the ends of genuineness variables. 188 00:33:55.980 --> 00:33:57.330 Julie Hogan: Or how the. 189 00:33:58.560 --> 00:34:07.830 Julie Hogan: Or how the JET tagging criteria sort of allow us to select these different jets and I apologize that we're so close to the end of the session already. 190 00:34:09.570 --> 00:34:16.830 Julie Hogan: I suppose, everyone is free to decide when they need to disconnect all of these exercises have solutions here available for you. 191 00:34:26.610 --> 00:34:41.370 Julie Hogan: For those that are able to stay on why don't we take about why don't we take if it's okay about let's say eight nine minutes until until 1835 and then I can share with you the solutions. 192 00:34:42.690 --> 00:34:44.040 Julie Hogan: All of the exercises. 193 00:34:44.040 --> 00:34:45.810 Julie Hogan: involved drawing things. 194 00:34:45.870 --> 00:34:50.250 Julie Hogan: From your from your output file that you got from running. 195 00:35:10.440 --> 00:35:16.080 Julie Hogan: i'm just to let everyone know I can't see the chat on either my devices, right now, so if. 196 00:35:19.080 --> 00:35:21.900 Julie Hogan: You need to ask a question go ahead and just ask. 197 00:35:32.370 --> 00:35:41.190 Dimitri Bourilkov: In the math you have to one of the methods was some and all the other rewards direct, why is this Meta analyzer. 198 00:35:43.440 --> 00:35:56.910 Julie Hogan: Sure, so the one that's the nomenclature i'm the one that's called some et is basically telling you that, rather than being in the negative vector some that's like the scale or something. 199 00:35:56.940 --> 00:35:59.310 Julie Hogan: that's like just just the energy some. 200 00:36:01.020 --> 00:36:18.030 Julie Hogan: So the ones that are called met PT met px met P, why those are talking about that vector that negative vector some over the event some et is is literally summing up a scale or quantity of the energy from all the deposits. 201 00:36:19.020 --> 00:36:21.780 Dimitri Bourilkov: Okay, I see thanks yep. 202 00:36:26.340 --> 00:36:26.970 Julie Hogan: Great question. 203 00:36:39.060 --> 00:36:43.500 Julie Hogan: So to do this first exercise live for you i'm going to choose to draw. 204 00:36:44.520 --> 00:36:45.870 Julie Hogan: The pruned mass. 205 00:36:47.520 --> 00:36:55.620 Julie Hogan: approved mass and I already know how to spell it because i've looked at the i've looked at the analyzer and seeing what branch I put. 206 00:37:03.240 --> 00:37:09.960 Julie Hogan: And so you can find a correlation here between masses and. 207 00:37:11.670 --> 00:37:15.060 Julie Hogan: can find a correlation between masses and. 208 00:37:17.550 --> 00:37:18.420 Julie Hogan: momentum. 209 00:37:21.150 --> 00:37:27.960 Julie Hogan: Where You see, this is the mass axis, you see the light works down here. 210 00:37:29.160 --> 00:37:42.060 Julie Hogan: You see right below 100 the WHO zones peering so we can actually observe them straight, you know straight away from the about 200 GB where we begin storing these large radius jets. 211 00:37:43.080 --> 00:38:00.450 Julie Hogan: But then we have a correlation appearing so in order to reconstruct top corks who they are within one of these jets, we need to make a higher need to make a higher criteria on our on our transverse momentum of maybe 400 or 500 GB. 212 00:38:04.380 --> 00:38:17.820 Julie Hogan: The solution pictures a little bit nicer because I did some roots of playing, and so you can really see the two peaks start to appear, where we have w bows ons and we have top works in this sample. 213 00:38:20.070 --> 00:38:24.210 Julie Hogan: So there's a correlation between massive momentum. 214 00:38:26.430 --> 00:38:27.690 Jose Monroy: Sorry, can I ask a question. 215 00:38:28.170 --> 00:38:29.370 Julie Hogan: Go ahead yeah. 216 00:38:31.530 --> 00:38:32.010 Jose Monroy: Can you hear me. 217 00:38:32.550 --> 00:38:32.790 yep. 218 00:38:33.870 --> 00:38:43.860 Jose Monroy: When you say Hughes the Web person or he's the CEO Boston you mean that the inside the JET the reason why awesome or sequel server or something like that. 219 00:38:44.580 --> 00:38:45.600 Julie Hogan: That is what I mean. 220 00:38:47.100 --> 00:38:50.010 Julie Hogan: So these are so so, for instance in this plot. 221 00:38:50.010 --> 00:38:56.670 Julie Hogan: Here around at are a bunch of w bows ons and since there since the momentum. 222 00:38:57.480 --> 00:39:07.110 Julie Hogan: Of this jet is so high, when the w decays the separation between the two forks is so small that they're both reconstructed within the same jet. 223 00:39:07.890 --> 00:39:20.700 Julie Hogan: So then, what we what we observe is a jet that represents the entire w goes on decay, or if we go up here a jet that contains an entire top work decay. 224 00:39:22.440 --> 00:39:24.150 Julie Hogan: And that's sort of what was pictured. 225 00:39:25.740 --> 00:39:36.390 Julie Hogan: up here, this is a jet that represents a top work decay, when you have let's say, probably, this one is probably the be quark and then these two are came from the w bowser. 226 00:39:39.600 --> 00:39:46.350 Jose Monroy: Of these lovely Boston, but these inside the JET is coming from somewhere else, I mean it's not from the from the main process right. 227 00:39:48.000 --> 00:39:59.940 Julie Hogan: But you okay in this picture, this is a top court decay, so that w comes from the decay of the top work but absolutely the w's can be produced in the interaction in the May in the primary. 228 00:40:02.280 --> 00:40:06.300 Julie Hogan: The primary collision or they could be the decay product of something more massive. 229 00:40:07.800 --> 00:40:15.060 Julie Hogan: In order for them to acquire this much momentum it's much more likely that they came from the decay of something very heavy. 230 00:40:19.380 --> 00:40:20.040 Jose Monroy: Okay, thank you. 231 00:40:20.760 --> 00:40:21.060 Julie Hogan: mm hmm. 232 00:40:30.540 --> 00:40:34.980 Julie Hogan: Okay, I mean I come down and do the second exercise for you have enough time. 233 00:40:36.390 --> 00:40:45.750 Julie Hogan: If you'll allow me a couple extra minutes here, we can also we could do the same type of command to look at the correlation between mass and. 234 00:40:46.530 --> 00:40:53.280 Julie Hogan: Between mass and and subject enos so i'm going to choose to use, I mean it stick with the w bows on idea here. 235 00:40:54.270 --> 00:41:13.470 Julie Hogan: And w bows ons use this TAO to to jet enos to sub dirtiness divided by one subject enos in order to make some distinguishing criteria, between those jets and other deaths so i'm going to plot pruned mass. 236 00:41:14.850 --> 00:41:19.620 Julie Hogan: versus this ratio TAO to over tell one. 237 00:41:20.910 --> 00:41:25.560 Julie Hogan: And I think i've got my axes oriented the way I want them, yes, I do. 238 00:41:28.350 --> 00:41:44.790 Julie Hogan: In this picture the different particles are showing up even more clearly, so if we look at the mass axis here are like quarks way down zero, but you can tell not literally falling from zero they tend to kind of peek in the second been here. 239 00:41:46.530 --> 00:41:57.150 Julie Hogan: We have a big blob of w bows on jets as would be expected from top work decays we have we have many high momentum w's being produced from these top works. 240 00:41:58.260 --> 00:42:12.180 Julie Hogan: And then, since this is a high mass TT bar production that we run we actually do have a lot of jets where the entire top work can be contained within within that jet and that's this blob up here. 241 00:42:13.950 --> 00:42:16.230 Julie Hogan: So you can see how you might. 242 00:42:17.670 --> 00:42:30.180 Julie Hogan: use my screen capture tool to show you, you can see how you might use these variables to grab your w goes on jets might do something like this. 243 00:42:31.410 --> 00:42:34.080 Julie Hogan: In order to grab the WHO zones. 244 00:42:35.250 --> 00:42:48.120 Julie Hogan: And that block where most of them live is between a massive like 65 to 105 in this pruned mass and values, less than. 245 00:42:49.470 --> 00:42:57.120 Julie Hogan: What there are various working point 0.6 0.45 you can choose where you want to draw a line over here. 246 00:42:57.630 --> 00:43:04.350 Julie Hogan: And if you take events, the left, you can see that you're rejecting most of the light works you're actually also. 247 00:43:04.740 --> 00:43:20.100 Julie Hogan: Rejecting most of the top works, so you really are looking for who zones that are that are contained within one jet, and so this is really useful if you're doing something like searching for a high mass what you're McCall it that decays to who zones. 248 00:43:23.580 --> 00:43:25.410 Julie Hogan: And I think in the solution here. 249 00:43:26.820 --> 00:43:30.210 Julie Hogan: I have that value that same plot, we were just looking at. 250 00:43:32.490 --> 00:43:41.400 Julie Hogan: And I also have the version for top part tagging where we have used the soft drop mass here on the on the vertical axis. 251 00:43:42.780 --> 00:43:55.710 Julie Hogan: And the three subjects penis divided by the two subjects penis which pushes the WHO zones way up here to large values and keeps the top works at smaller values. 252 00:43:57.690 --> 00:44:09.570 Julie Hogan: So, looking at plots like that that you can compare them to these tables and see how the selection criteria for these what we call boosted objects are motivated. 253 00:44:14.550 --> 00:44:24.270 Julie Hogan: So to close out with this last with this last little exercise this gives you some of the you know you can refer back to this and play around with it a little bit more later. 254 00:44:24.930 --> 00:44:46.050 Julie Hogan: Or, as your interest guides you this gives you some of the variable names for reference in your poetry, and you can you can just kind of look and see how you might clean up your selection of w's this would be w bows on selection the bet the tightest one here is in the red. 255 00:44:47.310 --> 00:44:55.380 Julie Hogan: Compared to the black where we have you know not done any selection at all, and we have a significant contamination of like quirks and top top works. 256 00:44:57.420 --> 00:45:04.110 Julie Hogan: Or you might you might similarly start to select your top Cork jets by using some of the top Burke criteria. 257 00:45:05.400 --> 00:45:13.080 Julie Hogan: So, starting with the 2015 data, all of this stuff became kind of stored by default to automatically in in the. 258 00:45:13.560 --> 00:45:29.370 Julie Hogan: Large radius jets for cms and so we've added that to this poet branch we didn't we didn't support it before for the 2011 to 2012 data, so now we kind of have the infrastructure for that, so that people can people can start to use it. 259 00:45:31.560 --> 00:45:31.920 Julie Hogan: All right. 260 00:45:33.210 --> 00:45:35.820 Julie Hogan: Little over but happy to answer any questions. 261 00:45:36.510 --> 00:45:47.730 Julie Hogan: And and throughout the rest of the workshop, if you want to look at the extra episodes on be tagging and the JET corrections and if you have any questions about that i'm certainly happy to ask them to answer them on matter most you can tag. 262 00:45:53.700 --> 00:45:54.480 Edgar Carrera: Thanks Julie. 263 00:45:57.270 --> 00:45:58.140 Edgar Carrera: And everyone. 264 00:46:00.600 --> 00:46:03.750 Edgar Carrera: You have questions final questions for Julie. 265 00:46:06.240 --> 00:46:08.040 Edgar Carrera: If not, I think we are. 266 00:46:09.150 --> 00:46:09.990 Edgar Carrera: very much. 267 00:46:11.310 --> 00:46:13.830 Edgar Carrera: Past the hour so. 268 00:46:16.110 --> 00:46:26.310 Edgar Carrera: we'll try to do a better tomorrow with the sound apologies for that, but I guess we'll see you, at the same time, tomorrow, thank you. 269 00:46:29.910 --> 00:46:30.270 Jose Monroy: Thank you. 270 00:46:30.750 --> 00:46:31.230 Ayman A. Al-Bataineh: Thank you.