path: root/2022/captions/emacsconf-2022-health--health-data-journaling-and-visualization-with-org-mode-and-gnuplot--david-otoole--answers.vtt

WEBVTT

00:00.000 --> 00:13.920
And we are live. Hello again, everyone. It's been a while. I've been on break for the last

00:13.920 --> 00:19.000
one hour. It feels like I've been really far. I'm with David. Hi, David. How are you doing?

00:19.000 --> 00:23.360
Hi, Leo. I'm good. Thank you for having me on.

00:23.360 --> 00:26.480
So David, you do have the pad open in front of you.

00:26.480 --> 00:27.480
That's right.

00:27.480 --> 00:32.800
Yes. I think Sasha is putting up the pad for us so that they don't have to look at my very

00:32.800 --> 00:38.520
tiny face on BBB. I'm very small. I'm trying to push the boundaries. David, if you want

00:38.520 --> 00:44.320
to take some questions, I'm sorry, I've got hells whispering in my ears at the same time.

00:44.320 --> 00:49.800
Give me just a second. David, can you maybe read the first question and try to answer

00:49.800 --> 00:50.800
them?

00:50.800 --> 00:58.640
Sure. Question one. Excuse me. Do you use this just for yourself or do you use this

00:58.640 --> 01:04.160
to discuss or show with doctors or health professionals? The answer is it is useful

01:04.160 --> 01:10.880
for me and it's useful for health care providers. So it's not I'm not just able to see connections

01:10.880 --> 01:15.440
myself, even if we're not seeing a connection. If I'm talking to someone, if I'm talking

01:15.440 --> 01:20.640
to a clinician, I have brought this in on a tablet. If it's an in-person thing or on

01:20.640 --> 01:26.040
the phone, if it's a video appointment with someone, just send it in an email. It's very

01:26.040 --> 01:33.160
convenient to just send the graph around as a screenshot. So it's definitely been really

01:33.160 --> 01:40.920
useful to because one thing that I touch on in the talk is that let's say that we're talking

01:40.920 --> 01:45.880
about depression, for example. Let's say, oh, well, how have you been sleeping this

01:45.880 --> 01:49.440
week? Well, if you had objective information, that's perfect. Go ahead.

01:49.440 --> 01:53.840
Oh, sorry. I was mispressing my button for production. You're good.

01:53.840 --> 01:59.960
Oh, it just caught a little clip of sound. I see. Okay. Sorry. Yes. So where was I? I

01:59.960 --> 02:05.960
would say, okay, so on the one hand, having objective information is good not only for

02:05.960 --> 02:09.360
you, but when it's time to speak to someone and give them a picture of what's going on,

02:09.360 --> 02:14.800
you don't have to generalize. Even if you're having a good day, generalizing can be really

02:14.800 --> 02:19.300
like, oh, well, how are you sleeping? Oh, well, this day I woke up that, but which way

02:19.300 --> 02:22.720
was that Wednesday or Thursday? And then I remember this friend came over, so it must

02:22.720 --> 02:29.000
have been Friday. That's not I've discovered that actually just first thing in the morning

02:29.000 --> 02:33.720
waking up recording. How many pieces of gum did I have yesterday? How much did I sleep?

02:33.720 --> 02:40.000
Okay. So someone else, this segues right into the next question. How do you input the health

02:40.000 --> 02:46.120
data? Semi-automated with Org Mode capture templates, copy paste, automated with a smart

02:46.120 --> 02:51.080
watch, and if this, then that. Tasker, Org Mode document to automatically add stuff like

02:51.080 --> 02:57.040
sleeping data. Which parts are and are not automated? Okay. So it turns out I'm using

02:57.040 --> 03:03.320
an Org Mode capture template to automate the insertion of all the little category tags

03:03.320 --> 03:10.600
in the property drawer, but not the values. I'm not using a smart watch or anything whatsoever.

03:10.600 --> 03:17.560
I am just, I look at the clock when I go to bed and I look at the bed when I wake, look

03:17.560 --> 03:21.480
at the clock when I wake up. The very first thing I do is turn on my computer and turn

03:21.480 --> 03:30.160
on Emacs and enter that data. So, you know, it would be really cool if we could get some

03:30.160 --> 03:35.120
kind of data, you know, because if it's all, you know, if it's be able to be massaged into

03:35.120 --> 03:40.760
Org Mode format, it can then go into Org GNU plot. Or maybe you could cut Org Mode out

03:40.760 --> 03:45.600
of the equation if you find a way to go directly from that data to either CSV, which Org Mode,

03:45.600 --> 03:50.760
I'm sorry, which GNU plot can read, or, you know, whether you have to wear a little script

03:50.760 --> 03:56.560
in between. And the next question, how do you track the various health statistics that

03:56.560 --> 04:02.720
you're gathering? The ones, you know, for example, nicotine is an estimate. I'm pretty

04:02.720 --> 04:08.460
sure it was six. It was pretty sure it was seven. You know what I mean? It's pretty close

04:08.460 --> 04:12.820
because I do count and I think I could at most be off by one or two. So it's not fatal

04:12.820 --> 04:16.980
to the enterprise of collecting data. The same thing with sleep. I look at the clock,

04:16.980 --> 04:23.240
but I only count half hours. I say, oh, I slept seven hours, you know, seven and a half.

04:23.240 --> 04:28.280
It's very much a back of the napkin thing, but it's precise enough overall that as you

04:28.280 --> 04:33.240
can see from the data in the talk, you can see that my sleep declined pretty steadily

04:33.240 --> 04:38.160
over the course of that. So even if some of the values are off a little bit, you're still

04:38.160 --> 04:44.060
seeing trends and you can still see connections. So it's not perfect. I would love to have

04:44.060 --> 04:49.680
some type of more automated data, but all right. So I'm going to move on to the next

04:49.680 --> 04:56.280
question. It's possible to download data from Apple Watch's health app. Is it easy enough

04:56.280 --> 05:07.000
to incorporate those.csv files into your implementation of gnuplot? Okay. So I think

05:07.000 --> 05:10.680
what it would involve if you wanted to use my template generator, it would involve making

05:10.680 --> 05:17.880
your template, but then modifying the gnuplot script to read from that file instead of from

05:17.880 --> 05:27.200
an org mode table. So I think that what you would want to do is select the, yeah, I mean,

05:27.200 --> 05:33.680
and if you want, you know, I'll be around in the IRC after and I might even be able

05:33.680 --> 05:38.200
to look at their documentation or something, just even just give a peek at it to give a

05:38.200 --> 05:39.640
better answer to this question.

05:39.640 --> 05:45.640
Cool. And just to remind everyone, so, sorry, let me try to put up, okay. I did crush Dev

05:45.640 --> 05:49.400
earlier when trying to switch window. Oh, okay. I managed to do it. Yes, I still have

05:49.400 --> 05:58.240
it in me. So let me, slightly later, we will also be opening the discussion. So this BBB

05:58.240 --> 06:01.520
instance in which we are currently, we will be opening it so that people can come in with

06:01.520 --> 06:06.160
questions and maybe you're considering the personal nature of this talk, you know, maybe

06:06.160 --> 06:11.520
try to be a little wary of sharing personal information, especially if the stream is live,

06:11.520 --> 06:17.000
but otherwise you can have a chat with David, provide to David you're still available in,

06:17.000 --> 06:21.880
you know, 10, 20 minutes and feel free to come in and ask questions because really that's

06:21.880 --> 06:26.480
what Emacs Conf is about. It's about getting in touch with the speakers and this is the

06:26.480 --> 06:30.200
opportunity that you have to have a one-on-one with them rather than an asynchronous one

06:30.200 --> 06:34.120
with a pad. Sorry for the interruption. I just felt I remember. We're probably going

06:34.120 --> 06:38.080
to open the chat in about five to 10 minutes. We still have a lot of questions and I know

06:38.080 --> 06:42.280
I'd prefer if David focused on those questions. So please go ahead, David.

06:42.280 --> 06:48.800
Thank you so much, Leo. And yes, I am available for after, so I'll stick around. Okay. So

06:48.800 --> 06:54.120
regarding the medication tracking, you only have the option to record missed or not. If

06:54.120 --> 06:58.280
one needs to take multiple medications throughout the day, how would you propose to track that

06:58.280 --> 07:05.240
within GNU plot or separate? Okay. Yeah, as it is, I don't record which medication because

07:05.240 --> 07:12.600
it's the one I miss is typically just the same one. So the deal is, is that you would

07:12.600 --> 07:19.660
probably want to, you could count it as either doses or milligrams. The problem with tracking

07:19.660 --> 07:27.660
as milligrams, and yes, you would want to use separate variables for those. And probably

07:27.660 --> 07:36.080
what would make sense is either to record some lines as doses, meaning that one, whatever

07:36.080 --> 07:43.040
that ends up being, if it's a tablet or if it's the proper dose. And then that way, if

07:43.040 --> 07:47.100
one medication is late and another is not, or whatever you need to record, they're on

07:47.100 --> 07:54.440
separate lines, separate columns of the CSV file or I'm sorry, not CSV file of your org

07:54.440 --> 08:04.540
mode, separate property drawer entries in your capture template. I'm trying to think.

08:04.540 --> 08:08.520
It would be a little bit more work, but you could track that. The problem with tracking

08:08.520 --> 08:13.760
milligrams is that some medications, you would have a large range on the chart because some

08:13.760 --> 08:18.320
medications are a few milligrams and some medications are lots and lots for a dose.

08:18.320 --> 08:27.720
So tracking it as single doses might make sense. All right. So how's the workflow? This

08:27.720 --> 08:32.120
is the next question here. How's the workflow when working on the canoe plot code? Can you

08:32.120 --> 08:38.560
control C, control C and the SVG output on the right is updated automatically? The answer

08:38.560 --> 08:46.200
is yes. You can use something called auto revert tail mode. So you hit meta X to get

08:46.200 --> 08:53.040
the command and then auto revert tail mode. And that will cause the SVG file to automatically

08:53.040 --> 08:58.840
update every time you hit control C, control C on the canoe plot file, provided that you're

08:58.840 --> 09:07.680
viewing the same file that the canoe plot is going to overwrite. Let's see. Question.

09:07.680 --> 09:13.000
How much time does it take to process the amount of data that you add inside canoe Emacs?

09:13.000 --> 09:21.360
So at this point I have three or four months of data and it takes to update the whole thing,

09:21.360 --> 09:27.360
like, you know, get capture all the entries. It takes like a second or two. And this is

09:27.360 --> 09:31.200
a pretty decent PC. So we're not talking about something that takes a long time each day

09:31.200 --> 09:39.840
I updated. It's just, it's done pretty quick. All right. Next question. Will indent guide

09:39.840 --> 09:45.120
behave well with YAML files for Helm? I don't know the answer to that. Indent guide is really

09:45.120 --> 09:52.720
nice and it seems to behave pretty reasonably, but I wouldn't be able to speak to that.

09:52.720 --> 09:59.280
Okay. Just a quick interruption. So we have opened the chats now for people to join us

09:59.280 --> 10:04.360
on BBB. So you can go to the talk page on Emacs. You go to David's talk, which is called

10:04.360 --> 10:08.760
health and you should be able to join the BBB. So if you have questions that you'd like

10:08.760 --> 10:13.800
to ask with your voice to David, feel free to join us. And David, in the meantime, you

10:13.800 --> 10:16.960
can reply to the last question you have.

10:16.960 --> 10:22.880
Okay. Thank you. Great. The last question is, have you noticed your behavior changing

10:22.880 --> 10:29.000
as a result of tracking your data? And the answer is yes. The accountability has given,

10:29.000 --> 10:36.160
it draws habits, good or bad habits. It illustrates them in sharp relief. All of a sudden it's

10:36.160 --> 10:44.440
very clear what's going on, what's not going on. And it illuminates a sense of fog because

10:44.440 --> 10:49.080
it's hard to generalize when your awareness is changing over time. It's hard to generalize

10:49.080 --> 10:53.560
about what's going on each day and having the objective information. Yes, I'd say it

10:53.560 --> 10:57.080
has changed my behavior in a good way.

10:57.080 --> 11:04.240
I mean, that's a very good thing, I suppose. The whole point of tracking so much stuff,

11:04.240 --> 11:07.680
tracking people, it's an interesting angle that you have because you're creating data

11:07.680 --> 11:12.320
on a lot of health aspects. And people usually when they talk about using Emacs to better

11:12.320 --> 11:16.000
their life, they talk about all, they talk about getting things done. They talk about

11:16.000 --> 11:22.560
all the usual suspects when it comes to task organization. But plotting over time is something

11:22.560 --> 11:27.120
that is technically possible and that people talk about, but I feel like you've gone way

11:27.120 --> 11:32.720
far into this particular topic. So well done you. And you feel as a result you have a particular

11:32.720 --> 11:40.800
insight into seeing the curves evolve basically over time. And I think it provides very interesting

11:40.800 --> 11:46.760
data, something that is more easily understandable than just a list of to-dos in an agenda over

11:46.760 --> 11:47.760
time.

11:47.760 --> 11:54.320
Yeah. And what's nice is that I still can have to-dos and notes in that same file that

11:54.320 --> 12:00.600
produces my chart and that has all the data in it. And I keep that in version control,

12:00.600 --> 12:05.040
which stays in a private repo. And what's nice about that is everything's backed up.

12:05.040 --> 12:11.920
You know what I mean? Pardon me. I'm going to have a sip of water here.

12:11.920 --> 12:19.680
Please do. I mean, I'm actually down to termite. It's the good floral for thermos. And I've

12:19.680 --> 12:23.120
down two since we started EmacsCon today.

12:23.120 --> 12:30.680
Oh, wow. Yeah. And so I lost my train of thought.

12:30.680 --> 12:36.840
Take your time to find it. We still have about eight minutes of Q&A. By the way, I'll remind

12:36.840 --> 12:41.200
people we have opened the BBB room. So if you want to join us and ask live questions

12:41.200 --> 12:56.120
to David, feel free to do so. I'm sorry. Could you repeat that? I was asking you if you found

12:56.120 --> 12:59.000
the train of thought again.

12:59.000 --> 13:08.480
Hold on. Hold on. The last question, have I noticed behavior changing? Yeah. Then we

13:08.480 --> 13:12.400
talked about, oh, I talked about the fact that you can have to do the notes and appointments

13:12.400 --> 13:18.080
and all that kind of stuff. Like, for example, one of my next. Well, oh, now I remember the

13:18.080 --> 13:22.560
point that I was going to get into. The point is that once I have this routine of getting

13:22.560 --> 13:28.600
up every single morning and first thing, putting in my information. That habit, I feel like

13:28.600 --> 13:33.200
build on. For example, I added a new variable the other day. You know, I can add I thought

13:33.200 --> 13:37.920
of something. I'm like, hey, I should track that in my data. It reminds me I say, hey,

13:37.920 --> 13:41.080
there's not enough green triangles. I need to get some exercise this week. You know,

13:41.080 --> 13:48.520
the weather, I haven't wanted to get outside because the weather's been so dismal. So it's

13:48.520 --> 13:58.240
absolutely been a game changer. It really has been. And it doesn't have to be. Oh, go

13:58.240 --> 14:04.880
ahead. Go ahead. No, no, no. I was going to say that it's really good to hear that software

14:04.880 --> 14:10.080
can have such a potent impact, just the ability to track your data over time. Because the

14:10.080 --> 14:14.480
thing is, I think whatever you're feeling, you know, whenever you have mental ailments

14:14.480 --> 14:20.320
or whatnot, it's really hard to to have a healthy relationship to the chronology of

14:20.320 --> 14:26.360
feeling better. And it's easy to get lost into the immediate sensation. But I feel like

14:26.360 --> 14:30.120
being able to track over time that, you know, this daily checking in that you do in the

14:30.120 --> 14:34.760
morning of keeping track of your health data. I think it's the first step that shows you

14:34.760 --> 14:42.800
over time that, oh, yes, I am actually feeling better. And so as a result, being able to

14:42.800 --> 14:50.440
visually see, oh, yes, I am actually doing better. It must be reassuring, I assume.

14:50.440 --> 14:55.060
It is. I think that's a really interesting turn of phrase you had there earlier when

14:55.060 --> 15:02.840
you said it's harder to have a healthy relationship to the chronology of getting better. I think

15:02.840 --> 15:10.160
those were your exact words there. You know, the idea that, you know, if you've got to

15:10.160 --> 15:16.080
manage something, if you've got to roll with the punches, then seeing what you've accomplished,

15:16.080 --> 15:21.440
seeing yourself bounce back, you know, and then being able to see a trend, you might

15:21.440 --> 15:26.920
not notice unless you're tracking per the clock that sleep has started to decline and

15:26.920 --> 15:33.880
that it's time to start keeping your eye out. And it is reassuring, and you're right about

15:33.880 --> 15:37.920
that, you know. It's not always easy to be in the mindset of saying, hey, these lines

15:37.920 --> 15:39.960
move and they'll move again.

15:39.960 --> 15:46.560
Yeah, and I feel also it's kind of like having a scientific take on your well-being because

15:46.560 --> 15:53.960
you are parameter, okay, really tough word to say for the Frenchman that I am, but parameterizing

15:53.960 --> 15:58.720
the elements of your life, such as sleep, and we already have sleep is already more

15:58.720 --> 16:03.960
or less of a parameter in our lives. You know, we know, oh, you should sleep at least eight

16:03.960 --> 16:08.820
hours. You should have one hour, 30 blocks of sleep all the time. So this is a well parameterized

16:08.820 --> 16:13.960
element of our lives. But other elements that you're doing or the conflation of different

16:13.960 --> 16:19.360
curves feels like it's pointing out, it's providing more data and being able to see

16:19.360 --> 16:23.920
the trend, which is a key word in your presentation in what you do. I think that's really good.

16:23.920 --> 16:28.800
Chronology and trends and being able to see what works. I just have this example crossing

16:28.800 --> 16:33.360
my mind. Some people are plagued with canker sores, which are really nasty stuff that you

16:33.360 --> 16:38.040
get in mouth. But more often than not, it's really hard to track why you're getting them.

16:38.040 --> 16:42.600
A lot of people say it's due to stress or to poor sleep or to deficiency in iron or

16:42.600 --> 16:47.780
whatnot, or you're toothpaste. You might have understood that I actually suffer from them.

16:47.780 --> 16:52.240
So I actually know a lot of the reasons. But being able to track health data like this

16:52.240 --> 16:57.560
feels like it would be able to correlate the appearance of canker sores with maybe poor

16:57.560 --> 17:03.240
sleep or maybe changing medication or maybe stress at work or stuff like this. I really

17:03.240 --> 17:06.240
like this stance and I think you're really onto something here.

17:06.240 --> 17:12.960
Yeah. And if people wanted to chat about that more in terms of putting people's heads together,

17:12.960 --> 17:19.120
because let's say someone who has a more sophisticated medication tracking need, maybe they come

17:19.120 --> 17:23.760
up with a solution and they share their config. There's one thing I didn't talk about in the

17:23.760 --> 17:31.580
talk, which is I had one line where I have a formula instead, because Canoe can plot

17:31.580 --> 17:36.760
all kinds of formulas. So I have it plot an overall trouble score, which I did not show

17:36.760 --> 17:45.520
in the video, but I had it assigned three points for every hour above or below the goal

17:45.520 --> 17:49.360
of eight hours. So too much would give me lots of points on the trouble and too little

17:49.360 --> 17:54.880
will give me lots of points. So the relation to the goal or whatever, if stress is high,

17:54.880 --> 18:00.920
then the overall, if the anxiety line is high, then the overall trouble line should get high.

18:00.920 --> 18:06.800
And what I found is that that trouble line really did track with my recollection of

18:06.800 --> 18:12.560
how the last few weeks had gone. And so that trouble line is, you can see when it's trending

18:12.560 --> 18:21.600
even if it's not super obvious from the other lines. And maybe I'll share that too.

18:21.600 --> 18:28.400
All right, David, we are about at the end of time for the Q&A. Thank you so much for

18:28.400 --> 18:33.000
answering all the questions. I think it was very valuable to not only have your long presentation,

18:33.000 --> 18:37.080
which was very detailed, but also have this Q&A where you got the chance to answer one,

18:37.080 --> 18:41.240
two, three, four, five, more than seven questions, which is great.

18:41.240 --> 18:46.960
That's fantastic. And thank you so much for joining in and for being here. And thank you

18:46.960 --> 18:51.720
for having me at the conference. It's really amazing to be part of the community. And I'm

18:51.720 --> 18:56.720
happy to stick around. Do you want me to drop to IRC or just keep answering questions in

18:56.720 --> 19:01.500
voice? I don't think they're going to go on the pad, are they?

19:01.500 --> 19:05.760
So actually what we can do, we can leave the BBB room standing for now. If people want

19:05.760 --> 19:10.240
to join, they can do so. Maybe Corwin will come back to close the room afterwards. But

19:10.240 --> 19:13.140
for now, you can stay in the room. And if people want to unmute themselves after we

19:13.140 --> 19:15.800
go off stream, that would be good.

19:15.800 --> 19:18.720
So we are about to go into the next talk in about 30 seconds. Let me just talk to production

19:18.720 --> 19:28.360
real quick. So, yes, we are going to go into the next talk pretty soon. You should still

19:28.360 --> 19:35.000
be able to hear me, I think. Nope, they cannot. Okay. Well, David, I leave you in the room.

19:35.000 --> 19:41.360
I'll be in touch within five minutes to see if people are not showing up.

19:41.360 --> 19:42.360
Thank you, Leo.

19:42.360 --> 19:48.160
Yes, sorry. I'm dealing with many things at the same time. So I'm going to stop the recording.

19:48.160 --> 19:52.680
No, I'm going to leave the recording on. Is there anyone in the room that want to unmute?

19:52.680 --> 19:56.760
We are not on the stream anymore now. Does anyone want to say something or talk with

19:56.760 --> 20:04.260
David? Or do we want to close the room?

20:04.260 --> 20:09.440
Even if it's just by line. Okay, cool. So what are we going to do then? David, thank

20:09.440 --> 20:14.080
you so much for your time. We're actually going to close the room right now. I'm going

20:14.080 --> 20:35.200
to stop the recording.