| --- Log opened Thu May 25 00:00:44 2017 | ||
| -!- TingMiao [uid229534@gateway/web/irccloud.com/x-yaintjrpptwrglzh] has quit [Quit: Connection closed for inactivity] | 02:23 | |
| @sukey | Pull Request #3812 "Add global fixture with binary label data" synchronized by MikeLing - https://github.com/shogun-toolbox/shogun/pull/3812 | 04:11 |
|---|---|---|
| @sukey | Pull Request #3812 "Add global fixture with binary label data" synchronized by MikeLing - https://github.com/shogun-toolbox/shogun/pull/3812 | 05:14 |
| -!- mikeling [uid89706@gateway/web/irccloud.com/x-qzzfxdjjjckqrmgv] has joined #shogun | 06:12 | |
| mikeling | wiking: ping | 06:12 |
| @wiking | mikeling, pong | 06:34 |
| @wiking | mikeling, i have commented on it | 06:34 |
| @wiking | if you have questions lemme know | 06:34 |
| mikeling | wiking: thank you | 06:34 |
| @wiking | mikeling, are you here? | 07:14 |
| mikeling | wiking: yep | 07:15 |
| mikeling | I'm working on the pr | 07:15 |
| @wiking | can i ask you a favour? can you go to this page https://github.com/shogun-toolbox/shogun/projects/1 and add a card to TODO (anything just a test) and a column? | 07:15 |
| mikeling | sure | 07:17 |
| mikeling | let me have a try :) | 07:17 |
| @wiking | thnx | 07:18 |
| mikeling | wiking: mmm, how could I add a card. I failed to find any button to do it | 07:18 |
| @wiking | :( | 07:19 |
| mikeling | it seems like I don't have that authority？ | 07:19 |
| @wiking | lisitsyn, just a reminder that the g2.8xlarge is still hanging around... it's fine until the 31st of may ;) | 07:36 |
| lisitsyn | wiking: I terminated it | 08:19 |
| lisitsyn | going to disappear | 08:19 |
| @wiking | k | 08:20 |
| @wiking | lisitsyn, you mean the g2.8xlarge? | 08:48 |
| @wiking | because that's still around | 08:48 |
| @wiking | g2.8xlarge | 08:49 |
| @wiking | sorry i-0939ca4da09151049 | 08:49 |
| lisitsyn | g2.8xlarge is still around yeah | 08:54 |
| lisitsyn | wiking: can we keep it for a bit more? | 08:55 |
| @wiking | sure till 31st is fine | 08:55 |
| lisitsyn | no I mean after :P | 08:55 |
| @wiking | mmm | 08:57 |
| @wiking | how much it drains a week? | 08:57 |
| lisitsyn | let me check | 08:57 |
| lisitsyn | not that much as p2 | 08:57 |
| lisitsyn | :D | 08:57 |
| lisitsyn | wiking: oh pretty much anyway. 2.8$ per hour | 08:58 |
| @wiking | 470 usd | 08:59 |
| @wiking | we have 5k for the whole year | 08:59 |
| lisitsyn | yeah will turn it off | 08:59 |
| -!- geektoni [~geektoni@93-34-234-212.ip52.fastwebnet.it] has joined #shogun | 10:24 | |
| -!- geektoni [~geektoni@93-34-234-212.ip52.fastwebnet.it] has quit [Quit: Leaving.] | 10:46 | |
| -!- iglesiasg [~iglesiasg@217.119.234.214] has joined #shogun | 11:31 | |
| -!- mode/#shogun [+o iglesiasg] by ChanServ | 11:31 | |
| -!- leagoetz [~leagoetz@eduroam-int-pat-8-238.ucl.ac.uk] has joined #shogun | 13:04 | |
| @iglesiasg | leagoetz: hola! | 13:05 |
| leagoetz | hey Fernando | 13:05 |
| @iglesiasg | leagoetz: I didn't get the chance to ask you last week before or after the chat with Ting, how did your concert at the Barbican go? | 13:06 |
| leagoetz | oh :) | 13:06 |
| leagoetz | it was really good. Super enthusiastic audience and we played well | 13:06 |
| @iglesiasg | cool | 13:07 |
| -!- HeikoS [~heiko@x55b47c72.dyn.telefonica.de] has joined #shogun | 13:07 | |
| -!- mode/#shogun [+o HeikoS] by ChanServ | 13:07 | |
| leagoetz | here's a review if you can be bothere: http://www.classicalsource.com/db_control/db_concert_review.php?id=14545 | 13:07 |
| @iglesiasg | do you play in an orchestra? | 13:07 |
| @iglesiasg | all right, I can probably find out in the article :D | 13:07 |
| leagoetz | yes, a symphony orchestra | 13:07 |
| leagoetz | in any case. | 13:08 |
| leagoetz | what's on our to do list before we talk to Ting? | 13:08 |
| leagoetz | or more general before the start of GSoC | 13:08 |
| leagoetz | the coding I mean | 13:08 |
| leagoetz | I should mention though that for Olivier the coding phase has already begun since he's away for some time during the coding phase so project shifted to earlier start | 13:09 |
| @iglesiasg | let me go through the doc she shared with us last week quickly again | 13:10 |
| leagoetz | we haven't heard back from her on any of the action points as far as I'm aware | 13:11 |
| @iglesiasg | I think tomorrow she will be prepared to tell us about that | 13:12 |
| leagoetz | it's quite a long list to do over a weekend - but I guess if she'd done some of it already then we would know? | 13:12 |
| @iglesiasg | but I'd say for the moment is to wait until we hear what Ting has come up with after the chat | 13:12 |
| leagoetz | sure | 13:13 |
| @iglesiasg | yeah, it is definitely a somewhat large list so maybe we shouldn't expect all the points to be addressed by tomorrow | 13:13 |
| leagoetz | Ok, how about I ping her to prepare an action plan (that finishes on Tuesday, where the coding begins) that we can discuss tomorrow | 13:14 |
| @iglesiasg | (I am referring to this "to do before GSoC" list) | 13:14 |
| leagoetz | yes I can see tha t;) | 13:14 |
| @iglesiasg | mmm I am not sure about an additional action plan; what would you expect from it? | 13:15 |
| leagoetz | I don't mean more fuzz | 13:15 |
| @iglesiasg | in particular, what would the action cover that is not already by this todo list and the gsoc schedule itself | 13:15 |
| leagoetz | basically I just want to hear how she is going to do all that before Tues | 13:15 |
| @iglesiasg | yeah, I agree | 13:16 |
| leagoetz | yes of course that list IS the action plan but so far there seems to be no action form it | 13:16 |
| @iglesiasg | I think tomorrow we can suggest that we would love to hear about progress without chasing | 13:16 |
| leagoetz | yup | 13:17 |
| @iglesiasg | it is true that gsoc has not officially started so it should not be a red flag at this moment | 13:17 |
| leagoetz | no, but it's about setting it up | 13:18 |
| @iglesiasg | but at the same time, since next week she's off for exams I expected a normal week already this one | 13:18 |
| leagoetz | like ski jumping | 13:18 |
| leagoetz | if you don't start with enough momentum | 13:18 |
| leagoetz | you won't get as far | 13:18 |
| @iglesiasg | definitely, I agree | 13:18 |
| @iglesiasg | but let's just wait for tomorrow and see what Ting brings us | 13:19 |
| leagoetz | check your email | 13:19 |
| @iglesiasg | I am expecting she will have some draft code and text for us | 13:19 |
| @iglesiasg | checking | 13:19 |
| leagoetz | yes, I'd like to ask her to send it now, so we get more than 2 secs to look at it before giving feedback | 13:20 |
| leagoetz | ok, I'll send an email to everyone | 13:22 |
| @iglesiasg | goedzo | 13:23 |
| leagoetz | done | 13:27 |
| leagoetz | let me know if anything else comes up, otherwise see you tomorrow | 13:27 |
| @iglesiasg | thanks for sending out the mail | 13:30 |
| @sukey | Pull Request #3812 "Add global fixture with binary label data" synchronized by MikeLing - https://github.com/shogun-toolbox/shogun/pull/3812 | 13:45 |
| -!- iglesiasg [~iglesiasg@217.119.234.214] has quit [Quit: leaving] | 14:05 | |
| -!- olinguyen [81615ad9@gateway/web/freenode/ip.129.97.90.217] has joined #shogun | 14:41 | |
| leagoetz | olinguyen: hey | 14:45 |
| leagoetz | how is everything going? | 14:45 |
| olinguyen | i'm doing good, how are you? | 14:45 |
| leagoetz | good! | 14:46 |
| leagoetz | where can I see some of it? | 14:46 |
| olinguyen | What I've been working on? https://github.com/olinguyen/gsoc2017-shogun-dataproject/ | 14:47 |
| leagoetz | would be good if you could send it to Heiko and me before we talk so we can take more time for a look and discuss better/ give you more valuable feedback | 14:47 |
| leagoetz | ah yes | 14:47 |
| leagoetz | I was about to say that exploring notebook could do with more visuals. But you already have that on your to do list :) | 14:52 |
| leagoetz | note though that you need to move on from preprocessing/exploring to building the logistic regression, SVM etc models, as this is your first week plan (remember the dates are shifted as you started the coding this week, not next week) | 14:54 |
| olinguyen | Yea, I hope to get some preliminary results by the end of this week | 14:55 |
| leagoetz | have you made any proof of concepts for these models yet? | 14:55 |
| olinguyen | Not really. But what would be a proof of concept for say logistic regression used in this case? | 14:56 |
| leagoetz | to implement one | 14:56 |
| leagoetz | but it doesn't have to be perfect yet | 14:57 |
| @HeikoS | hey hey | 14:58 |
| @HeikoS | Ill join in a sec | 14:58 |
| @HeikoS | olinguyen: leagoetz | 14:58 |
| @HeikoS | hey, so whats up? | 14:59 |
| @HeikoS | how are things going olinguyen? | 14:59 |
| leagoetz | the sun 8D | 14:59 |
| @HeikoS | haha | 14:59 |
| olinguyen | Not much sun here yet :P | 15:00 |
| olinguyen | But i'm doing good! What about you? | 15:00 |
| @HeikoS | yeah good here, greetings from Berlin | 15:00 |
| @HeikoS | olinguyen: would you mind very briefly walking us through the things you did before Tuesday, and what happened since then? | 15:01 |
| olinguyen | Sure | 15:01 |
| @HeikoS | any surprises, any interesting output? | 15:01 |
| @HeikoS | any problems? | 15:01 |
| @HeikoS | and whats up for the next few days | 15:02 |
| olinguyen | Before Tuesday, I finished the training to obtain access for the MIMIC database, prepared and go the setup running. I also spent some time reading the MIMIC paper and some of its documentation | 15:02 |
| olinguyen | From Tuesday, I explored the dataset to obtain general population statistics and individual patient information, what type of data is included in the dataset | 15:03 |
| leagoetz | wait, I thought that was supposed to be pre-GSoC? | 15:03 |
| olinguyen | Part of it was done pre-GSoC but I carried over a little on Tuesday | 15:04 |
| leagoetz | ok, sure | 15:04 |
| olinguyen | I knew that the data was large, but there was more content that I thought (types of measurements, quantity of measurements per patient, etc) | 15:04 |
| @HeikoS | olinguyen: cool, so whats the outcome of the stuff? | 15:04 |
| leagoetz | sure just thinking about where we might be short on time | 15:04 |
| olinguyen | So i'm currently at the preprocessing phase, a bit more needs to done before I build the models | 15:05 |
| olinguyen | But this should be within the next day | 15:05 |
| -!- iglesiasg [~iglesiasg@217.119.234.214] has joined #shogun | 15:05 | |
| -!- mode/#shogun [+o iglesiasg] by ChanServ | 15:05 | |
| @HeikoS | olinguyen: anything particular thats worth mentioning there? | 15:06 |
| @HeikoS | any problems? | 15:06 |
| olinguyen | There was a little ramp up time for my to pick up some SQL stuff that I didn't know. Some issues I had, but solved, were getting the docker running with Shogun and the MIMIC database | 15:06 |
| leagoetz | yes, important point - if there's something cumbersome/ peculiar about the data/ etc that also should go into your end-of-week text | 15:06 |
| leagoetz | and eventually blog | 15:07 |
| @HeikoS | olinguyen: so you have a running docker with shogun and the mimic? | 15:07 |
| olinguyen | yes | 15:07 |
| @HeikoS | cool | 15:07 |
| leagoetz | so write it down (and while you're at it send me in email ;)) | 15:07 |
| @HeikoS | leagoetz made a good point there | 15:07 |
| leagoetz | (just trying to help you be efficient :) | 15:08 |
| @HeikoS | olinguyen: so I am still kind of waiting for some summary email to send to gunnar, could you compile something there? | 15:08 |
| olinguyen | Sure! | 15:08 |
| @HeikoS | olinguyen: cool thanks | 15:08 |
| leagoetz | great! | 15:08 |
| @HeikoS | olinguyen: the other thing that we discussed were some poc drafts | 15:08 |
| @HeikoS | (proof of concept) | 15:08 |
| @HeikoS | of the models you want to run | 15:08 |
| olinguyen | Right | 15:08 |
| @HeikoS | initially those were supposed to happen before the start, to reduce the number of surprises | 15:08 |
| @HeikoS | but I guess we should do those soon | 15:09 |
| @HeikoS | maybe as part of the email stuff as well? you think that is feasible? | 15:09 |
| leagoetz | like I said - doesn't have to be perfect | 15:09 |
| olinguyen | By PoC, do you mean building basic models with preliminary results? | 15:09 |
| @HeikoS | olinguyen: exactly | 15:10 |
| olinguyen | Got it | 15:10 |
| @HeikoS | emphasis on basic | 15:10 |
| olinguyen | Ok | 15:10 |
| olinguyen | yes, i think it's feasible | 15:10 |
| @HeikoS | cool | 15:10 |
| @HeikoS | olinguyen: you realise we discussed the same things last week? ;) | 15:10 |
| olinguyen | Yes, sorry :( | 15:10 |
| @HeikoS | olinguyen: no worries, we dont want to be too pushy you know, we mostly want to help you get stuff done | 15:11 |
| leagoetz | +1 | 15:11 |
| olinguyen | I understand | 15:11 |
| @HeikoS | What is the plot in the notebook? just found that guy | 15:11 |
| @HeikoS | first 24hrs | 15:12 |
| olinguyen | It's a plot for a single patient with 4 physiological measurements | 15:12 |
| olinguyen | measurements were made during the first ICU stay | 15:13 |
| leagoetz | I would have a couple of suggestions for the plot, but I don't want you to get lost in details there, so ask me at a later point when you have some spare time maybe? | 15:13 |
| olinguyen | Sure | 15:13 |
| @HeikoS | I am mostly curious what it shows us | 15:13 |
| @HeikoS | or what one can learn from it | 15:13 |
| @HeikoS | olinguyen: anything comments on this one? Or is it more a placeholder showing that the data was extracted correctly and we have access to those time-series | 15:14 |
| olinguyen | We can see that the measurements are sampled irregular times. Perhaps adding the outcomes e.g. whether the patient died or not, or how long he stayed in relation to this plot could be more informative | 15:15 |
| leagoetz | (OK, while we're at it - would be good if you could plot the quantities relative to their "normal" value. Only medics will be able to tell you whether NIBP of 60 is good, or bad etc) | 15:15 |
| @HeikoS | I was about to suggest the same, some bounds for those values would be useful | 15:16 |
| olinguyen | Good point | 15:16 |
| @HeikoS | olinguyen: those can also be useful to construct features for prediciton later | 15:16 |
| @HeikoS | olinguyen: what do you think, which of your 3 prediction tasks would this data be useful for? | 15:16 |
| olinguyen | I thought of heart rate, but that is likely to be fluctuate alot and be different from the normal behaviour for most patients | 15:17 |
| @HeikoS | olinguyen: not sure I follow that | 15:18 |
| @HeikoS | I meant the 3 objective predictions you have, mortality, hospital readmission, length of stay | 15:18 |
| leagoetz | do you have the death cause for these patients? you might have to subset ... I'm thinking people who died of heart failure, heart rate should be informative | 15:19 |
| olinguyen | Right. (Forgot a bit what I said earlier) | 15:19 |
| leagoetz | but this may not be the case for people who die from something lung-related (-> blood oxygen?!?) | 15:20 |
| olinguyen | True. Though i think HR still is a good indicator of the patients overall health, which are useful for length of stay | 15:20 |
| leagoetz | can you actually show your plots to a medic? | 15:20 |
| leagoetz | I'm just guessing here, it would be good for you to discuss your results with someone who knows a bit about the subject | 15:21 |
| olinguyen | I don't have anyone in my surroundings but I could look around | 15:21 |
| leagoetz | btw I need to leave at 14:30 | 15:21 |
| @wiking | HeikoS, pingu | 15:22 |
| @HeikoS | wiking: jojo | 15:22 |
| @HeikoS | olinguyen: gunnars group! | 15:22 |
| leagoetz | but you can keep discussing with HeikoS afterwards | 15:22 |
| @HeikoS | so make that part of the collection of things that we send to him | 15:22 |
| leagoetz | YES! | 15:22 |
| @wiking | HeikoS, ok when can we chit-chat | 15:23 |
| @wiking | ? | 15:23 |
| olinguyen | Alright | 15:23 |
| @HeikoS | wiking: yep | 15:26 |
| @HeikoS | olinguyen: one last thing from my side, I will send you an email with the daily/weekly GSoC duties soon (we will share them jointly with all students in a meeting, but since you started early it will be via email) | 15:27 |
| @HeikoS | olinguyen: from my side, I would wait for some written summaries from your side then | 15:27 |
| olinguyen | ok | 15:27 |
| @HeikoS | olinguyen: cool let me know if anything comes up | 15:29 |
| olinguyen | Sure! | 15:29 |
| @HeikoS | wiking: see email | 15:29 |
| @HeikoS | wiking: jojo | 15:29 |
| @HeikoS | wiking: I am terribly busy as usual | 15:30 |
| @HeikoS | but maybe lets talk :D | 15:30 |
| @HeikoS | call? | 15:30 |
| olinguyen | HeikoS, leagoetz: I'll draft an email for gunnar within the next couple days | 15:30 |
| olinguyen | Would you like to wrap this up? | 15:30 |
| @HeikoS | olinguyen: wrap up? | 15:30 |
| leagoetz | maybe put a day on it | 15:30 |
| @HeikoS | olinguyen: write it in a way that it can be public | 15:31 |
| leagoetz | but yes, otherwise just let me know if something comes up | 15:31 |
| @HeikoS | olinguyen: (as we will ask you to make it public in the email :D ) | 15:31 |
| leagoetz | need to run, sorry | 15:31 |
| olinguyen | Ok sure | 15:31 |
| @HeikoS | olinguyen: so define your goals, say what data you have and show it, say what the next steps are | 15:31 |
| olinguyen | Got it | 15:31 |
| leagoetz | cool, soon we'll have the first model for that data! :D | 15:32 |
| olinguyen | *by wrapping up, i meant conclude the meeting | 15:32 |
| leagoetz | well you're supposed to hang out in irc anyways while coding ;P | 15:32 |
| olinguyen | Yes, ofc :). I just didn't want to left hanging | 15:32 |
| leagoetz | sure. that's it from me | 15:33 |
| leagoetz | bye! | 15:33 |
| leagoetz | speak soon! | 15:33 |
| @HeikoS | olinguyen: ah yeah sure | 15:33 |
| olinguyen | See you! | 15:33 |
| @HeikoS | leagoetz: bye! | 15:33 |
| @HeikoS | olinguyen: i have my irc open most of the time in case something comes up | 15:33 |
| @wiking | HeikoS, ok | 15:33 |
| @HeikoS | wiking: shall I facetime you? | 15:33 |
| olinguyen | HeikoS: great, thanks! | 15:33 |
| -!- leagoetz [~leagoetz@eduroam-int-pat-8-238.ucl.ac.uk] has quit [Remote host closed the connection] | 15:42 | |
| -!- leagoetz [~leagoetz@eduroam-int-pat-8-238.ucl.ac.uk] has joined #shogun | 15:43 | |
| -!- leagoetz_ [~leagoetz@eduroam-int-pat-8-238.ucl.ac.uk] has joined #shogun | 15:44 | |
| -!- leagoetz [~leagoetz@eduroam-int-pat-8-238.ucl.ac.uk] has quit [Ping timeout: 255 seconds] | 15:47 | |
| -!- leagoetz_ [~leagoetz@eduroam-int-pat-8-238.ucl.ac.uk] has quit [Remote host closed the connection] | 15:48 | |
| -!- geektoni [~geektoni@93-34-234-212.ip52.fastwebnet.it] has joined #shogun | 15:49 | |
| -!- leagoetz [~leagoetz@eduroam-int-pat-8-238.ucl.ac.uk] has joined #shogun | 15:49 | |
| @wiking | HeikoS, i can you cannot | 15:49 |
| @wiking | HeikoS, lemme know when i can ping u | 15:49 |
| @HeikoS | ? | 15:49 |
| @HeikoS | wiking: shall we facetime now? | 15:50 |
| @wiking | yes | 15:50 |
| @HeikoS | ok ill call | 15:50 |
| @wiking | i'll call u | 15:50 |
| -!- leagoetz_ [~leagoetz@eduroam-int-pat-8-238.ucl.ac.uk] has joined #shogun | 15:50 | |
| @wiking | HeikoS, i think you cannot | 15:50 |
| @wiking | dont ask | 15:50 |
| @wiking | it's a bit complicated | 15:51 |
| @iglesiasg | xD | 15:51 |
| -!- leagoetz [~leagoetz@eduroam-int-pat-8-238.ucl.ac.uk] has quit [Ping timeout: 246 seconds] | 15:53 | |
| -!- leagoetz_ [~leagoetz@eduroam-int-pat-8-238.ucl.ac.uk] has quit [Remote host closed the connection] | 16:17 | |
| -!- leagoetz [~leagoetz@eduroam-int-pat-8-238.ucl.ac.uk] has joined #shogun | 16:18 | |
| -!- leagoetz [~leagoetz@eduroam-int-pat-8-238.ucl.ac.uk] has quit [Ping timeout: 240 seconds] | 16:22 | |
| -!- leagoetz [~leagoetz@eduroam-int-pat-8-238.ucl.ac.uk] has joined #shogun | 16:36 | |
| -!- leagoetz [~leagoetz@eduroam-int-pat-8-238.ucl.ac.uk] has quit [Remote host closed the connection] | 16:39 | |
| -!- leagoetz [~leagoetz@eduroam-int-pat-8-238.ucl.ac.uk] has joined #shogun | 16:41 | |
| @wiking | HeikoS, fyi gatsby is back | 16:48 |
| @wiking | oh no i have to fix something there | 16:48 |
| @wiking | just a sec | 16:48 |
| -!- leagoetz [~leagoetz@eduroam-int-pat-8-238.ucl.ac.uk] has quit [Remote host closed the connection] | 16:49 | |
| -!- leagoetz [~leagoetz@eduroam-int-pat-8-238.ucl.ac.uk] has joined #shogun | 17:02 | |
| -!- geektoni [~geektoni@93-34-234-212.ip52.fastwebnet.it] has left #shogun [] | 17:05 | |
| -!- iglesiasg [~iglesiasg@217.119.234.214] has quit [Quit: leaving] | 17:07 | |
| -!- leagoetz [~leagoetz@eduroam-int-pat-8-238.ucl.ac.uk] has quit [Remote host closed the connection] | 17:08 | |
| mikeling | wiking: ping | 17:31 |
| @wiking | pong | 17:35 |
| @wiking | i'll check asap | 17:35 |
| mikeling | thank u :D | 17:37 |
| -!- leagoetz [~leagoetz@eduroam-int-pat-8-238.ucl.ac.uk] has joined #shogun | 17:50 | |
| lisitsyn | uhmm HeikoS how do you see this | 18:03 |
| @HeikoS | ? | 18:04 |
| lisitsyn | sklearn | 18:04 |
| @HeikoS | I mean the api is just really nice | 18:04 |
| @HeikoS | the methods they have | 18:04 |
| @HeikoS | the abstraction | 18:04 |
| lisitsyn | we have more or less the same | 18:04 |
| @HeikoS | and very consistent across the methods | 18:04 |
| lisitsyn | what exactly? | 18:04 |
| @HeikoS | yes but messy as fuck | 18:04 |
| @HeikoS | like the preprocessors | 18:04 |
| @HeikoS | they have all these nice things as "score" "inverse_transform" "transform" "inplace"-options | 18:05 |
| @HeikoS | very intuitive | 18:05 |
| lisitsyn | I don't agree everything is that nice in sklearn ;P | 18:05 |
| lisitsyn | I don't like superoverloaded constructors of methods | 18:05 |
| lisitsyn | and some things are not flexible enough | 18:05 |
| lisitsyn | like kernel='rbf' | 18:06 |
| lisitsyn | and if you want custom you do | 18:06 |
| lisitsyn | kernel='custom' and some crazy callback | 18:06 |
| @HeikoS | yep thats not good | 18:06 |
| lisitsyn | I know what you mean but I am a bit against pure copying | 18:07 |
| @HeikoS | but there is lots of good stuff | 18:07 |
| lisitsyn | yes copy good stuff don't copy bad | 18:07 |
| @HeikoS | and that is what I am talking about | 18:07 |
| @HeikoS | sure | 18:07 |
| @HeikoS | but nice API would be nice | 18:07 |
| @HeikoS | and we have superoverloaded constructors atm | 18:07 |
| lisitsyn | :P | 18:07 |
| @HeikoS | they will go with tags i iknow | 18:07 |
| @HeikoS | yep thats all I am saying | 18:07 |
| lisitsyn | it is nice to be rich and healthy | 18:07 |
| lisitsyn | and young | 18:07 |
| lisitsyn | HeikoS: what would be the only thing you could steal from sklearn? | 18:08 |
| @HeikoS | preprocessor base class methods | 18:08 |
| lisitsyn | HeikoS: could you elaborate? | 18:10 |
| @HeikoS | lisitsyn: sorry Im busy, I just thought this. Shall we sit down and write API for the new interfaces at some point soon? | 18:12 |
| lisitsyn | sure | 18:12 |
| lisitsyn | for both | 18:12 |
| -!- leagoetz [~leagoetz@eduroam-int-pat-8-238.ucl.ac.uk] has quit [Remote host closed the connection] | 18:56 | |
| -!- leagoetz [~leagoetz@eduroam-int-pat-8-238.ucl.ac.uk] has joined #shogun | 18:57 | |
| @HeikoS | wiking: still there? | 19:02 |
| -!- leagoetz [~leagoetz@eduroam-int-pat-8-238.ucl.ac.uk] has quit [Ping timeout: 260 seconds] | 19:02 | |
| @wiking | HeikoS, yes | 19:19 |
| -!- HeikoS [~heiko@x55b47c72.dyn.telefonica.de] has quit [Ping timeout: 260 seconds] | 19:29 | |
| -!- HeikoS [~heiko@x55b47c72.dyn.telefonica.de] has joined #shogun | 19:30 | |
| -!- mode/#shogun [+o HeikoS] by ChanServ | 19:30 | |
| -!- HeikoS [~heiko@x55b47c72.dyn.telefonica.de] has quit [Quit: Leaving.] | 19:36 | |
| -!- HeikoS [~heiko@x55b47c72.dyn.telefonica.de] has joined #shogun | 19:45 | |
| -!- mode/#shogun [+o HeikoS] by ChanServ | 19:45 | |
| -!- HeikoS [~heiko@x55b47c72.dyn.telefonica.de] has quit [Client Quit] | 19:48 | |
| -!- HeikoS [~heiko@x55b47c72.dyn.telefonica.de] has joined #shogun | 19:48 | |
| -!- mode/#shogun [+o HeikoS] by ChanServ | 19:48 | |
| -!- HeikoS [~heiko@x55b47c72.dyn.telefonica.de] has quit [Ping timeout: 260 seconds] | 19:53 | |
| -!- mikeling [uid89706@gateway/web/irccloud.com/x-qzzfxdjjjckqrmgv] has quit [Quit: Connection closed for inactivity] | 20:03 | |
| -!- leagoetz [~leagoetz@eduroam-int-pat-8-238.ucl.ac.uk] has joined #shogun | 20:31 | |
| leagoetz | lisitsyn: it is nice to be rich and healthy and young ... haha | 20:32 |
| lisitsyn | ain't it? | 20:33 |
| leagoetz | who am I to know, I'm only the latter two | 20:33 |
| CaBa | "only". reason enough to feel lucky, don't you think? :) | 20:58 |
| leagoetz | I'm a pretty happy person yes :) | 21:12 |
| leagoetz | happy to be lucky | 21:12 |
| leagoetz | whatever | 21:12 |
| -!- leagoetz [~leagoetz@eduroam-int-pat-8-238.ucl.ac.uk] has quit [] | 21:12 | |
| --- Log closed Fri May 26 00:00:45 2017 | ||
Generated by irclog2html.py 2.10.0 by Marius Gedminas - find it at mg.pov.lt!