 Okay, which microphone am I talking into into the computer in the laptop. Okay. Okay. This is, in a sense, quite a quite a nice follow on from what Barbara was was talking about Barbara was really focusing on you know one of the fundamental issues which is in scientific data sets you've got columns, and somehow you've got to describe, you know, what that column of numbers is actually describing. And what she is screen. Yeah, keep going. Well and she was describing saying that they've they've attempted to develop a model to sort of break out into its components or into it into its pieces what a particular number would describe and of course one of those. He's still finding me. Here we are. One of the key bits of the description of a column in a set of data very often is the unit of measure by which the number that you see in a column is scaled. You know, you'd think I've been barking on about this for 10 or more years now already you would think this would be a solved problem units of measure, almost the simplest part of the puzzle compared with biology and chemistry and and all of those things. You know, because it's so simple perhaps it's actually remains a relatively unresolved issue. And, you know, basically we have this situation where everyone who's developing a project to collect some particularly quantitative data, you know they almost always start out with building a list of the units of measure which they're going to be using for their data. And the codes that occur in that list of units of measure and not necessarily consistent, not all people using units of measure, you know know about uppercase and lowercase and whether to use a slash or quotes or, you know, the traditional units of measure a type set in beautiful ways with italics and Greek letters and all of those kinds of things. And I'll show you one very bizarre one, which is very, very widely used, which where the units of measure don't resemble anything you've ever seen. So, you know, one of the biggest problems is that a lot of data is thrown away with a rather tenuous connection to the metadata about what the numbers in the columns and the tables mean. So, you know, one of the tenuous connections is what what are the units of measure here. What what is actually the scale for this number here. And then there's a kind of a social problem here which is, who is who is it who runs units of measure well, there's a whole community out there called metrologists who worry about the ninth decimal place in the conversion factors between units of measure definitions are met in some applications that ninth decimal place really does matter. Most of the time it doesn't be better if we at least knew within maybe two decimal places what the the the units of measure is that we're doing. But the metrologists of course they own this space, but they're actually pretty shit at telling people what it is they're talking about in a way that makes it easy to reuse. You know, and here's the sort of prime example the organization that manages the the international system units the SI is the Bureau of International Weights and Measures based in Paris, and you go and look up to find the definitions of weights of units of measure there and what you get you get web pages and PDFs. Okay, and the web pages and PDFs are got beautifully typeset and all of that but there's no link to the definitions of individual units of measure that you can get back a description in a way that you can use in an application in a digital application. So this is, you know, the my point for that the metrologists know what they're doing but they don't know how to tell users how to make sense of it. There's a few projects out there, which have claimed or had a go at doing a general solution to units of measure. One of the key ones being what one of the most successful ones I think is this thing called QDT quantities units dimensions and types. It's, it's a quite a nice technical solution. There's a uri per unit of measure, however they are cached statically so if you come up with a unit of measure that isn't in their list, then you've got to go through a process of getting it added to their list can take as little as a week, but it's not, you know, the next minute. So they have, I hope you can read that the these are compact you arise that unit colon stands for an HTTP blah blah blah thing, but the name of the unit of measure is something that's kind of possible to a human. So what they've got there actually they have a fairly strict rule for how to do this so it's pretty predictable what you're getting out. And they've got a whole bunch of interesting units of measure in there. Currently, nearly 2000 instances of units of measure which is why this is one of the best ones out there, including things like ACA and Australian at some stage early in their piece they decided they were going to do units of currency and I think that was a bad idea and I've told them that. But so there's not a lot of attention paid to that but but but you know they do have quite a big set. What do you get if you request one of these descriptions of a unit of measure. I'm showing you there. There's a description of cubic inch per minute in represented as a JSON which anyone who's a web developer these days would at least be able to pass that into a data structure, and you can more or less read the kind of information in there. Line for there is the conversion multiplier to convert cubic inches per minute into the SI unit for that quantity type which of course would be cubic meters per second. Everyone knew that right. But you can see a whole bunch of other information in there which is quite useful in being able to use this and make sense of it there's a dimension vector I'll draw your attention to which is, which basically is classics of dimensional analysis. The dimensions per minute is got is length to the power of three and time to the power of minus one in that in that identifier there for the dimension vector. And then there's a whole bunch of other stuff including down the bottom there just your attention to a set of codes using the you come system which I'll talk about a little bit later. In science space. Oh, I'm sorry that that project was originally developed to the project was funded by NASA, but is now a kind of standalone project and that's one of the weaknesses of it, which is that essentially depending on trying to get funding to support their system from subscribers. And, and the people who are running it are all retired or semi retired people of my vintage or older, who know what they're doing but what happens if any of those falls off the perch. So this is another significant system which is widely used in one of the communities which I think is represented in this room which is the oceanography community, the NERC vocabulary service based on British Oceanographic Data Center. They've got a whole set of control vocabulary is online in the NERC vocabulary services, several hundred. This one is their units of measure and there's a couple of hundred units of measure in here. The URIs, or the codes for them in the ID column over the left there, some of them make sense some of them you can't understand, you know what units it's referring to you'd have to look it up, but but still it's, you know, a reasonably good system. So if you do reference one of the, those, those definitions is an example femtomoles per liter, you get a web page for that. And you can also download a machine readable version again I'm showing the adjacent representation of that. There's less information in this than there is in the QUDT definitions for the same things. In most cases there's actually a link across to the QUDT ones so that you can cross reference. There's already an existing community that's expecting to use the P06 codes, but there is the mapping across to the QUDT ones I'm not sure if I chose one of these here. But for most of them there is. So, it doesn't have the conversion factors built in here, for example. And so there's a limited it's just a pure, almost a pure SCOS representation. Here's an incredibly important list of units of measure put out by the UNC fact which is basically units of measure for for trade matters this is, you know, what, what is used for commerce in the world. You know, I was saying before there's some odd codes used that those are the codes for the units of measure in the UN ECE system, and you'll see some of them kind of make sense like the decimeter and centimeter, but who would have known that E96 is degrees per second. So, you know, you have to look them up you have to know them, they are used very, very widely in in the standardization and commerce. There's some beautiful concepts in their systems here. You put container tank truck, catch shot, a stick military thing. So, you know, the, the, the range of things that need units of measure is, is pretty beautiful. I mentioned earlier you come you come is a system that was originally developed in clinical clinical health applications are still basically owned by that community. So the big advantage of you come is that you can type it on a keyboard without using the shift key actually you may have to use the shift key to get some of the parentheses but it's, it's, it's, you know, it's basically no funny characters seven bit asking. It defines an algorithm to it has a set of atomic codes, and there's algorithm to how to combine them. The combinations look like the codes that you know and love. So they look like, you know, m slash s for meters per second and here's a bunch of examples of the slightly weird weird well some of the weird ones, you know per kilogram is slash kg. And the ones which are anyone who's got a scientific background would probably be able to guess. You come. Also, there are some quite nice tooling that comes along with it, including a validator and a converter. So here's a little web app that so I put in one inch of mercury as a measure of pressure. So let's put it to hectopascals which is another unit of measure for pressure. Bang, there it is one inch of mercury is equivalent to 33 and we're nearly 34 hectopascals, you know, perhaps a slightly more. Well, and there's some pretty weird ones in there. Yard acres, which would be a measure of volume as you can see 56 yard acres would be 207,000 cubic meters. So that's, but it's all based on just those little codes and I showed this one particularly to to indicate, you know that as well as all the standard si ones which are the ones you expect the so called conventional units the ones which are used in, you know for historical or cultural purposes of all that square brackets square brackets around them, but otherwise kind of makes sense. So, given the existence of a few of these different systems here there's been a team actually associated with the oboe foundry ontology system, who put together a little tool at units of measurement dot org, which basically looks up those systems for you. So, here's an example if you, and it uses the you come codes as the keys so you type in the you come code, and it's, it's, it shows you what the equivalent would be in the in nbs and p zero six in QDT and then a couple of other systems as well. So, there, there, there is an online mapping tool already available. So this is, you know, partly a mess and partly good news, I think there are some systems which are there. Not everybody knows about them, everybody's using them. So, you know, recognizing it as an issue co data established a task group on trying to calm down this. These these troubled waters. A few years ago, the task group was initiated in 2018 I'm on it. It's basically trying to come up with some, some systems recommendations in order to sort of bring some order to this chaos. One of the things we have to recognize and we have to accept is there will never just be one system. There are there's a lot of history here, communities have got their own needs they've got their own systems which are built around some of these existing services and systems. Getting everyone to throw them away in favor of, you know, the perfect system ain't going to happen. But if more of the, you know, new projects and new groups were aware that these existing systems were available, maybe they wouldn't have to do their own would realize they don't have to do their own. And the code the drum digital representation units of measure group has put out we had a paper in nature earlier this year. I think my paper with my biggest number of co authors 17 co authors on that some of whom I'm not actually sure what they did but it was important to have them as listed as co authors when you're making a pitch for for some kind of social change. But basically the appeal is, you know, if we don't resolve this problem then we, we've got a risk of having data sets out there which are not going to be as usable as they should be in the future. And in particular there's a liaison now for with the, this is the parent body of the International Bureau of Weights and Measures this sort of the board, CIPM, which has a task group on the digital SI. I've had a little bit to do with facts to chalk who I think is talking next is going is, I think is is is on this this committee. I think I briefed them or I briefed one of their sister committees in in in CIPM a few weeks ago on the existing sort of state of the art there, and believe it or not there is still argument about what the word quantity and quantity value and quantity and a unit actually mean. So, you know, most of us would recognize that the, this is the, the basic information model which QDT uses, and most of us would be able to make sense of this particularly if you've shown an example like this which says, Okay, this mass is a quantity kind mass. 1.54 kilograms that's an example of an individual mass so you need basically to understand all three of those bits of information in order for that number to make sense it's a mass it's it's in kilograms, and the value is 1.54 I should skip this slide sorry. So I'm just going to wrap up by saying, I think what I've tried to show is there are some good units of measurement lists that are in existence. Some of them are provided with on request, you know, in real time machine readable representations with the conversion factors. There are tools and crosswalks available for some of them. And the CIPM is liaising now with the digerati partly facilitated by codator through the through the drum project. So, I'll finish there.