 We're going to open the file, it's just not working. Okay, so if somebody is asking them, when hasn't converted, I will read on the calculation on mine. So one thing that I need to, I can try to walk you through the tutorial in the beginning and then there is a good chance that you are not going to be able to finish this tutorial in one hour. It may take a bit longer, although we try to reduce the parameters, the values that should take just a few minutes to calculate. So hopefully everybody knows by now how to get to this file. Yeah, so if you haven't done it, you can clone it and then you need to go into this directory where you have date five, PM1, EPW. So I'm already there in my case. Oh, sorry, not anymore. Okay, I need to reconnect. A few minutes while I reconnect. So this tutorial really takes you from the beginning. So initially you will need to get everything from quantum expressory. So the first step will be to run an ACF calculation in order to generate the charge density. So this is hopefully you are familiar with quantum express, but right now if you are not, doesn't matter, you only need to run the job. And then this input file is to perform a final calculation. So in this case, we are using a very small mesh of only two by two by two. So okay, I'm gonna first just do this. How long is gonna take? I'm sorry, I should have done that. And then I'm also gonna submit the phonons. Okay, I am leaving it on the screen so everybody can see. So I don't know, but somebody in the forum was saying that it takes very long. So you can see it's very fast for me. It's already in the six mode out of nine. But again, this is a very small Q mesh. So this calculations are not by any far, anyway converged, yeah? So it's just to be able to get something running. So in this case, for a two by two by two mesh, we should only have to calculate four Q points, yeah? So with every each of these four Q points, we are going to have nine modes since MGP2, there are three items in the unit set. I hope this is my last Q point. So right now you can see that we have this directory, pH, yeah? For people that are not familiar with Quantum Espresso. And for each Q point, we will have, so for the first point, the gamma point, this is the deformation potential file that I was talking about. So this is the DDSS. And then for all the other points, three points, you will have the same type of file inside this sub directory. So if you go here in the Q2, that is you'll find another DDSS file, yeah? So this is MGP2 DDSS. So in order to provide this information to EPW, we need to collect all this files with the deformation potential and also with the dynamical matrices. And this can be done very easily using this writing script. So yeah, so you just need to go to pay this line. If you do it, you should ask, sorry, you didn't copy the right file. Do it again. And then it's going to ask for the name. So in this case, the seed file is MGP2. And then if you will see here that there is a new directory that was created, the same directory. And in this, we have collected the DBSCF, so the potentials, yeah, this MGP2 underscore Q1, and also the dynamical matrices files. And also here in the last directory, I think we have the better files. So here we have the pattern, dynamical matrices with the pattern. So these are basically things that are needed for EPW calculations. So now the first step before running EPW is that we need to have some information for the Vanie or 90. So in order to do this, we need to run one non-sense, non-search consistent calculation on the dash, sorry, fine chain mesh that you are going to use for the vanillization. So in this case, again, the mesh is not, the dense is a six by six by six, uniform chain mesh. So this is the same you have already done things like this. No, I already answered that question. Yeah. And again, I'm going to go into the copy and do it on the screen so everybody can see. Right now, I'm only doing the NNICF, this should be very fast. Because they see this, they see this, then. Just to remove it, okay, let me close it. Oh, you can leave it, you can leave it. You see, I see, okay. So it's not important. Oh, because I didn't want it for time. I didn't want it for my speed, I didn't want it. Thank you. So now we have all information from one to MESPRESO that we need in IPvW. Now we have calculated the idle values. This will be required for the Vanille library and we have also found the phone. So we have basically, now we can finally do IPvW evaluation. So in this case, as I said, there are quite a lot of input flags. So it will follow the same style as in Vanille MESPRESO. So this is the prefix that we also use in your calculation for, guys, reduce a bit the sound there. Stepan is not hearing me or I can. So you need to use the same prefix that you use in Vanille MESPRESO calculations. And one important thing that you shouldn't forget, you need to provide a path to where you have the DBSCF files. So this here is the path to your, this directory where you remember I saved the theme matrices and the DBSCF so the potentials. Because if you keep here the wrong path, the code is going to crash. I think it provides an error message, but that's something you need to keep in mind. And here we are just setting the code run when you double do medical form calculations. Then, okay, these are the historical reasons that we started and let's take it more with the same. And this flag here, it just says the very first time when you do it, you are going to write the electron form matrix element in the binary representation on the file. And this is useful if you want to start your calculation. And so in the beginning, this needs to be true and the reason is false. Some of these flags are going to change over in the following. I'm gonna have a new version of the code release, but for now it's going to work with the current version. And here we have to say that the code will basically do a vanearization. So again, if this is the first run of IPW, vanear, this flag needs to be true. If you already have the vanearized electron form matrix elements, this needs to be turned off to false or otherwise you are going to repeat again the vanearization. Yeah, nothing will happen that you will spend time on the vanearization. And then sub here, this is really the number of vanear functions that you are going to use. Yeah, this may not be the same. And in most cases, it's not the same as full number of fans. Yeah, that's the same as in the vanearization. And then the way that we provide information to the vanear library. So we have a few flags that we can read them, but then a lot of flags we use for this. So these are the projections. There are other things we use. Okay, I don't see that it's in this file, but we have an array for W data. And there you can provide the pack and other things, other parameters for vanearization. Now, at the stick, you already seen the name of this input flag somewhere. So this is the Fermi window around the Fermi level. So whatever value I put here, this is going to be plus and minus. Yeah, so the Fermi window effectively is 0.4 say 0.2 below and 0.2 above the Fermi window. And this is the smearing that it's used for the data functions for the elements. And when we have this delta E and K minus, yeah, this is the smearing that we use. These are just additional things you want to plot. One important thing currently for when you solve the equation is that you need, since the solving this particular equation is very expensive and usually you cannot do it in the same run, you will need to restart. We write the electron form on the accelerant on five. And this are the electron form on the accelerant on the fine mesh, yeah, whatever mesh we use. Again, these are flags just to turn on the specific activity calculation to be understood through. And then here I specify that I'm doing anisotropic. We solve them anisotrophically. There is also, you can also solve them anisotrophically. And then it will be anisotrophic. Here, Elima means that you are going to solve them on this imaginary sequence axis. And then Epa there, it means that you are going to also perform an analytic continuation to a real axis using probably approximations. And this is the number of iterations in the self-consistency for solving the equations. That's a threshold. This WS cut is the cut in the matzubara frequency. As I said, usually you take it about at least five times the largest form of frequency. And this is this new star parameter that the semi-empty will show with attention. And here you would specify, here I will give you this example, right now we are just doing two, but how many temperatures you need to solve this equation. Yeah, so here I already know what result I need to guess. Yes, so here I am just specifying 10 and 20. But in general, you need to have some idea about what critical temperature and then the code gives a simple estimate using a methamphetamine program. I didn't provide the amount of it for that. And here we just give the mesh, yeah? So this is our original force meshes. This mesh is the exact same mesh that you, on which you did the NSEF calculation, the one for the vanearization. And this is the mesh on which the form and calculation was done, yeah? So this is the Q mesh. You remember it was two Q, here it was 60. And this are the fine meshes on which you are gonna calculate the superconductive properties, yeah? So in this case, again, the transport one thing that I forgot to mention, this is done consistently. You need to have the Q mesh and K mesh are commensurate. Otherwise the code is going to stop. It's not gonna work. And if you have this flag through, it means that it's going to only use the K points in the reusable brilliance zone, yeah? So this is going to significantly speed up your calculation because if the system has a high symmetry, it will reduce the number of K points quite well considered. So now we can start this run. Yes, I'm gonna copy this line and now I'm going to start this calculation. You can see there are questions in the forum, in the chat. Yeah, so initially this part is calculating, you can see it's even writing, it's calculating the first the MN and then the MNN, and then it did the vanyarization. So now it has finished with the vanyarization and now this part is calculating the electron format accelerants from the core, yeah? So these are reducible to be on four Q points, that was the number of irreducible Q points that we had on these two Qs, right? It's working, somebody was saying that for them the calculation of points, yeah, I don't know what you mean, you saw it, but you're not sure what you mean. And it's fine. So I think the break is at quarter to four. Quarter to four, really? Yes, it's a break, but then you have more, you know, you have about from four to five. But this is somewhere. Okay, okay. Yes. So three quarters of an hour or more. This one probably made a question, you know, in the literature of the discussion of the S way, the P way, this is also sweet, even though we call this, what we have to get on pi and sigma. On this way, it's all within the S way. Okay, but it's all within the S way. Because whatever is beyond this way, it's not PCS, it's not the right platform, you know? Okay, I'll just. So all of this is a break, okay, thank you. Yeah, no, this is a good point, I forgot to mention. So, thanks, see you, see you later, school. Okay, so I think the calculation is finished here. So I see that the number of files has been creating some are equal for plotting. But the, okay, so here in this document, I'm not gonna read everything, but basically it explains what every file contains. So as I said, the electron-formal matrix elements are returned to file, and this is, like right now we need only two temperatures, yeah, but in principle, you may need to do more temperatures and you want to restart. So then it's useful to have this information available so you don't have to restart the calculation again from a long stretch, yeah? Because this is the part that takes quite a lot to evaluate the electron-formal matrix elements on the fine mesh. And one thing that you need to keep in mind that if you restart, you will need to use the exact same number of processors that we used when you created this file. And the reason for this that you'll see, you'll say, if we go in this directory, there are as many files as a number of processors. So here I'm gonna use NPRan4, so I'll use four processors, so there will be four files. So if you restart the calculation, you need to restart it on port, otherwise the code is not going to work. And this other file, EGNV, this contains the idle values, and this FREQ contains the final frequency, and this IKMAP, some information about the symmetry, the other folding of K plus Q on K. So now we've already seen this, that's how it was calculated. That's what we are not gonna do right now, but in principle, this Q file can be used with Vesta or with Fermi Surfer to produce the gap on the Fermi surface. If you remember in my presentation, I had a lot that somewhere did a long time ago with Vesta, I think nowadays it's much easier to do it with Fermi Surfer. For Fermi Surfer, you are only going to use the files, but if you want to do it with Vesta, there is some explanation at the end, and I think you can find still on YouTube an old video recorded by somewhere, explaining all the steps of folding with Vesta. And I am just rewriting here the equations. So I think one useful point I like usually to look at is this lambda and K, because this can give you, you can do this without calculating the area, solving the area of the equation. So we can get an idea about an isotropy in your system. So for instance, if I look, I think there is a plot here, I look at this plot for NGP2. These are similar plots that you should get today. I'm very overwhelmed, but these are some nicer plots. If you look here, this is the, this lambda and K. So we can see that there are lambda concentrating two regions, which are well separated. So that's kind of an indication that most likely your superconductor will have two gaps. If you see such a large, let's say, split between in lambda, that's an indication. And that can give you an idea if you need to calculate, let's say an isotropic properties, or you can just go a lot and do an isotropic calculation, which can be done basically, you only need this area of the spectral function. You can even take this from quantum expressions, for instance, and you can run it in serial and that's basically a couple of minutes. And you can get the gap, that's really, really true. Other files that, so as I said, if you don't have any idea about what value you need for the temperature, the code in the very beginning provides an estimate based on Elandine's formula. So we can, and this can be done, and I think this is done, even if you don't, I think this is also done, even if you just ride the, look at the lambda. I don't think you need to run any Li-Archberg calculation at all. So you'll get this, you'll get an idea of your DC and that can be an idea of what values you should provide to the code. And as I said, if you want to do an analytic estimation, my recommendation is to use Pade because it's very cheap. This will require a lot of memory and it's, and basically need to solve the equations iteratively again on the real axis. And again here, so these are parts of this that are plotted. In this case, there are a number of files that will be provided. So this MG2 lambda K pairs, for instance, I said this is one file that I find very useful. We'll give you this spot here. This one lambda pairs basically gives you this anisotropic element. And basically this is just a measure of how many elements from the whole pool will have, let's say, lambda values of about 0.5. And then this last file, this is just the Li-Archberg spectral function, isotropic Li-Archberg spectral function. And this is the integrated lambda. And when you estimate this critical temperature based on the island line formula, all you need is this value of lambda. So with all this, I think that's basically the first run. Now, once you have some idea about the superconductivity, you can also upload the superconducting gap. So for this, you are going to have a bunch of files in this name where XX is going to be the temperature. Yeah, so if I look here in the code, you will see that I have these files in GB2. And here we need a calculation of two temperatures, so 10 and 20. In fact, that's the certain files that you will get now. These are just for the restarting options. And if you are going to do the analytic continuation with Fade, you are going to have an extra file that will have this subscript Fade. If you will do it with analytic continuation, you will have even more files. So you need to be careful because these files are pretty large, so you shouldn't write all of them. And if you are going to plot, I'm not gonna do it now, but if you are going to plot, let's say your file, this file, it's 10 KD. In this current round that you just have, you should generate a plot like this. You see that it's not very well defined there if they need to spread almost this band, or this band, if you are going to do it on a denser mesh, you will see that the tools are more isolated and this one comes from the pie and this is gonna come from the signal bands, yeah, the electronic signal bands. And so this is as a function of imaginary frequency. You see that there is no structure and that's why if you want to get some structure and some structure representation after that, like for instance, to calculate the quasi-particle density of states in the split conducting state, we need to go to real space. And if you then use the slides, you can create, you go here, you will get something like this, a flow like this and this is so you can compare it against experiments that show channeling data of quasi-particle density of states. And you see here that you'll have two peaks, one that corresponds to the lower gap and the one that is bigger than the higher gap in GB2. So I'm just jumping between different plots but just to get an idea of what kind of quantities you are getting here. So as I said, in order to do a star calculation, if you want to either, maybe you want to plot or to get the gap at some temperature for plotting to write these cube files because in general, I wouldn't recommend writing them during the round, since they are very large and then you are gonna write a file at every temperature, you may probably are gonna be interested just in the low temperature range, at your lowest temperature. Then you can just do it by re-reading or doing a restart calculation. So what you will have to do in that case, you see right now I have a dual restart, I already have calculate so I'm putting in the force because this doesn't matter, it's not gonna do anything because I put all the southern flags post. And then vanuatization is also forced and then you need to have this extra flag for reading if you do a mag underscore two and the code basically, and you need to have the file for the first sensor for the job providing this list. So the first, so in other words, you need to have this file that I'm just highlighting here, emug underscore an iso and the temperature that is the first temperature in this list. If you don't have this, the code is going to stop saying that it's missing the file because it provided you not read. If you want to start without starting from a previous standard, you don't have this slab, that's okay. And then if you start the calculation with a gas for the gas based on the BCS formula and it will do the temperature that you already have it here is the first standard. And then the calculations are the same exactly as before. So, we'll do this now, let me just copy. And you can see we have added a few more temperatures. If there is a temperature larger than the gap, the code is going to stop before it is not gonna do it. Yeah, so this is it, I mean, these are normal. So now, of course, this is way off yonder and this plot is like no means to reverse but this takes hours and you will need a few, yeah, maybe even a few hundred courses or so. There is no way we can run this shop here. But what we are plotting here from this mge, emug and iso file, this basically extracts the from all the spots here. It's basically the gap in the zero frequency. Yeah, so it's basically this first thing. And that's what you were super far back here. So you do that at every temperature and then you get this plot of the gap as a function of t. And as I said, when the gap becomes zero, this is a critical temperature in your superconductor. Now, these calculations, the anisotropic thing that should warn you if you have a superconductor with very low TC will be very, very expensive. Like most of the time, you cannot go and calculate below like 2 K. And the reason for that is that lower the TC, more matibata frequencies you are gonna have for a specific cup. Yeah, because of the expression. I don't think I have here this thing, but i omega j is basically two n plus one t. So smaller the t, denser the match for a specific cup. So it becomes very hard to calculate. You will need a lot of memory if your temperature is very low. On the other hand, it's nice that in the hydrides temperature is very high. So these calculations are relatively cheap because they have a TCO, most of them are above 100 Kelvin. And here, just a few more other things that I said that you can close. The superconducting quasi-particle density of states, once you have done the analytic continuation, you basically, as the code evaluates this formula, and that's this file contains again, the quasi-dose at every temperature. And if you want optional, okay, we just sign in it, you can, I'm not sure if here we can, we should do this, but in principle, you can try to increase your piece and you can do the calculation and see how the convergence is going. But I wouldn't recommend to do it on four processors here. They can also change the Coulomb parameter, as I said, you can look a bit more. And I'm not gonna go to this example, but as I said, you can also run the isotropic case. This can be done very simply on a single processor once you have this file for the Eliash perspective function. If it doesn't have to be produced necessarily with EPW, you can produce this file with quantum espresso on the, let's say, on the mesh, you can have a calculator in the format. And if you use that, that file, you can run the isotropic calculations. And in this case, for MGB2, of course, if you are going to calculate the gap with the isotropic thing, you only get one gap there. You don't get that MGB2 is a two-gap superconductor. And here, just a few more things about the starting options. I already said that this requires to use the same number of cores as the original run, that's very important. And for the restart, you need basically these files. You need, if you start, so there are different options at which point you are going to restart. If you didn't have enough time, maybe allocated time to finish writing the electro-format experiments and the fine mesh, then in that case, you can restart the calculation. And for this, you are going to need the requirement this file that I highlighted here. I think this is old already. We don't think we have the MGB2, yeah. So we need all these files that are provided here. And then for this, that's the input, the flags that are going to be set up in EPW. So first of all, you will need to read the EPMOD VP. So this means that you read the electro-formatic elements in the valier representation. And it's going to start recalculating just the ones on the fine mesh that have not been calculated in our previous class. And on the other hand, if you already have the electro-formatic elements on fine grid, if you have this, you can restart by using this file and this will be when you go directly to your sub-activity calculation. And this, directly, you will have the following files that I already showed. And you will have, besides the electro-formatic sub-files, you are going to have these three additional files. Yeah, and that's the way that you need to set up the flexibility. We only need to make these falls. You are not going to write them anymore. In fact, even if you have them true, if the code finds that they exist, it's not really going to recalculate them. And finally, additional information that is provided here. As I said, how to plot the sub-conducting gap using Vesta. You see that there are a number of steps that need to be done. And if you do it, that's a nice plot because there is also a tutorial on how to adjust and how to color it. And from this point on, this information can also be found on our website, but this is a description of the variables related to the sub-conductivity calculations. Questions? Questions? Not all right. So this is the gap question for each band and which group one? Yes. Do you want to give me one? Yes. And this is actually the question. Okay, so then you can have the part where you analyze which has the highest probability in the gap function, right? The first one? Yes. The first one? Yes. The first one? Yes. The first one? Yes. The second one? Yes. The first one? Yes. The second one? Yes. The first one. Yes. The second one? Yes. The second one? Yes. The second one? Yes. Yes, I see. You see that there are some things that you can think of. basically that also brings me to the next question, like the principle, this is now only calculated for the bands that we created by the functions, right? So, but this also means that we could afterwards do some kind of projection on the planning functions, right? You could, we wanted to do that. But like in principle, right now, we see it already because the Fermi service pockets some sheets that are probably formed by a specific offer. Yes, yes, yes. Because I mean, yeah, we kind of knew that some of the packets in the case, we didn't have anything. So, yeah. So, it was interesting, yeah. It was something that I saw. I should try to provide this. I mean, when we do it, I mean, we also calculated based on the Unfortunately, we don't get that. But we also, we also calculated another part of the green Spanish, right? There are several ways we do it. This paper would be in plastic, even if you do it like this, especially for relations. But this is us, like, and then the other one is just like spin bearing, like, yeah, like some spin, like the issues that they do with the bearing and would be nice to compare them to see, like, just, like, the data on the screen. So, yeah, there are some, there are also some, there are also some money. Exactly, so that's the point. Yeah. Yes, that's something that, yeah, something that I would like to point out. Yeah. Yeah. Yeah. Yeah. Yeah. Yeah. And then you know, you know, why am I thinking about that? Yeah. Yeah. Yeah. But I guess what it says is how to say correct, like write it down. Yeah. Yeah. Yeah. Yeah. Yeah. Yeah. Yeah. Yeah. Yeah. Yeah. Yeah. Yeah. Yeah. Yeah. Yeah. Yeah.