ChanServ changed the topic of #mlpack to: "mlpack: a fast, flexible machine learning library :: We don't always respond instantly, but we will respond; please be patient :: Logs at http://www.mlpack.org/irc/
mindlifter has quit [Remote host closed the connection]
mindlifter has joined #mlpack
b100s has joined #mlpack
mindlifter has quit [Remote host closed the connection]
b100s has quit [Remote host closed the connection]
mindlifter has joined #mlpack
rascul4 has joined #mlpack
rascul4 has quit [Remote host closed the connection]
mindlifter has quit [Remote host closed the connection]
mindlifter has joined #mlpack
Andreas_K[m] has joined #mlpack
Andreas_K[m] has quit [Remote host closed the connection]
mindlifter has quit [Remote host closed the connection]
mindlifter has joined #mlpack
drakonis has joined #mlpack
drakonis has quit [Killed (Sigyn (Spam is off topic on freenode.))]
MLNoob22 has joined #mlpack
MLNoob22 has quit [Client Quit]
Guest52790 has joined #mlpack
Guest52790 has quit [Remote host closed the connection]
mindlifter has quit [Remote host closed the connection]
mindlifter has joined #mlpack
UmarJ has quit [Read error: Connection reset by peer]
mindlifter has quit [Remote host closed the connection]
mindlifter has joined #mlpack
jrgilman2 has joined #mlpack
jrgilman2 has quit [Remote host closed the connection]
ImQ009 has joined #mlpack
mindlifter has quit [Remote host closed the connection]
mindlifter has joined #mlpack
mindlifter has quit [Remote host closed the connection]
mindlifter has joined #mlpack
sajfer22 has joined #mlpack
sajfer22 has quit [Remote host closed the connection]
honningkanon28 has joined #mlpack
honningkanon28 has quit [Remote host closed the connection]
mindlifter has quit [Remote host closed the connection]
mindlifter has joined #mlpack
mindlifter has quit [Remote host closed the connection]
mindlifter has joined #mlpack
ImQ009_ has joined #mlpack
ImQ009 has quit [Ping timeout: 260 seconds]
ImQ009_ has quit [Ping timeout: 260 seconds]
mindlifter has quit [Remote host closed the connection]
mindlifter has joined #mlpack
hsmath has joined #mlpack
hsmath has quit [Remote host closed the connection]
ImQ009 has joined #mlpack
StasM[m] has joined #mlpack
StasM[m] has quit [Remote host closed the connection]
LarchOye19 has joined #mlpack
LarchOye19 has quit [Remote host closed the connection]
ImQ009 has quit [Ping timeout: 245 seconds]
mindlifter has quit [Remote host closed the connection]
< jonpsy[m]> @zoq Updates on the remote machine?
mindlifter has joined #mlpack
Guillaume[m]1 has joined #mlpack
Guillaume[m]1 has quit [Remote host closed the connection]
< zoq> jonpsy[m]: Commented on the gdoc let me know if you run into any issues.
< jonpsy[m]> okay it looks like the entire thing is empty?
< jonpsy[m]> it looks like i need sudo access to install things
< jonpsy[m]> can i ask you to setup ensmallen, mlpack and xeus cling?
< jonpsy[m]> * do you mind to setup ensmallen, mlpack and xeus cling?
< jonpsy[m]> .......aand armadillo :)
< zoq> You can also use docker on the machine, that way you have root as well.
< zoq> docker pull mlpack/jenkins-amd64-debian
< zoq> That one has everything you need I think
< zoq> besides xeus cling
< swaingotnochill[> zoq I have some doubts regarding the linear regression implementation.
< swaingotnochill[> I trained the model on trainSet, trainLabels
< swaingotnochill[> Then I predicted the responses on testSet as output
< swaingotnochill[> So, I am confused in Compute Error function. Does, it automatically uses the outputs of testSet to calculate error with testLabels.
< swaingotnochill[> (edited) ... with testLabels. => ... with testLabels?
< swaingotnochill[> Also, is there any way to convert arma:: rowvec into normal vector?
Chaz6 has joined #mlpack
Chaz6 has quit [Remote host closed the connection]
< jonpsy[m]> <zoq "That one has everything you need"> I guess that way I won't need ssh, no..?
< zoq> swaingotnochill[: Not sure I get the question, you have to pass the ground-truth labels and the predicted responses to the compute error function to get the L2 squared error, internally the compute error function uses the trained linear regression model.
< jonpsy[m]> <swaingotnochill[ "Also, is there any way to conver"> See "arma::conv_to<>::from() "
< zoq> with normal vector you mean arma::vec?
< zoq> Right arma::conv_to<arma::vec>::from(...); should do the trick.
< jonpsy[m]> <jonpsy[m] "I guess that way I won't need ss"> Abt this?
< swaingotnochill[> I think the ComputeError takes a matrix and responses as input
< swaingotnochill[> By normal vector means, will I able to use .begin(), .end() operators , coz I was first trying to calculate MSE
< zoq> So std::vector? It does work for std::vector as well.
< jonpsy[m]> I think rowvec has begin end
< zoq> Yes, Mat, Col, and Row have iterators as well
< rcurtin[m]> swaingotnochill: if you're trying to compute MSE, I might suggest writing it as a single operation using Armadillo functions... `arma::mean((vec1 - vec2) ^ 2)` I think should work
< zoq> swaingotnochill[: Ahh, I see what you mean, in this case you just pass the testData, the trained model will do the prediction internnaly and compare it with the passed ground-truth.
< rcurtin[m]> in general, the more high-level operations you can use, the better job Armadillo will do of producing efficient code
< swaingotnochill[> zoq yaah...ComputeError calculates L2 loss by its own
< swaingotnochill[> <rcurtin[m] "swaingotnochill: if you're tryin"> Yaah..Will try using armadillo functions...
danrik13 has joined #mlpack
danrik13 has quit [Remote host closed the connection]
ponyofdeath5 has joined #mlpack
ponyofdeath5 has quit [Remote host closed the connection]
ImQ009 has joined #mlpack
Dieterbe13 has joined #mlpack
Dieterbe13 has quit [Remote host closed the connection]
JamesNZ has joined #mlpack
JamesNZ has quit [Remote host closed the connection]
mindlifter has quit [Remote host closed the connection]
mindlifter has joined #mlpack
mindlifter has quit [Remote host closed the connection]
mindlifter has joined #mlpack
LiveWire has joined #mlpack
LiveWire has quit [Remote host closed the connection]
mindlifter has quit [Remote host closed the connection]
mindlifter has joined #mlpack
< DavidportlouisDa> @zoq I tried to one hot encode my data in python binding and then this happend ![error_image](https://drive.google.com/file/d/1Xiy7Aa5VjD_3JlVBhEueXMeotm-82EUa/view?usp=sharing)
< DavidportlouisDa> * @zoq I tried to one hot encode my data in python binding and then this happend ![error_image](https://drive.google.com/file/d/1Xiy7Aa5VjD_3JlVBhEueXMeotm-82EUa/view)
< DavidportlouisDa> * @zoq I tried to one hot encode my data in python binding and then this happend ![error_image](https://i.imgur.com/OlXuLfA.png)
< DavidportlouisDa> * @zoq I tried to one hot encode my data in python binding and then this happend
< DavidportlouisDa> ![error_image](https://i.imgur.com/OlXuLfA.png)
< zoq> DavidportlouisDa: Do you have X at dimension 6,8 as well?
< zoq> Is that numerical?
< DavidportlouisDa> No it's string (region name & type)
< jeffin143[m]> Currently we support only numeric data type
< jeffin143[m]> We need to find out a way to , take input as string
< jeffin143[m]> I mean how do we store that matrix in the binding ???
< zoq> Right, that is for the executable/python binding.
< zoq> Do you think you can do the encoding in pure python as well?
< jeffin143[m]> zoq: I didn't get you ??
< zoq> So we don't have to rely on the preprocess_one_hot_encoding.
< zoq> I was just suggesting to drop the one_hot_encoding in the notebook, since categorical data is not supported
< zoq> at least for the python API
< DavidportlouisDa> ok
< zoq> C++ should work, @jeffin143[m] correct me if I'm wrong.
< jeffin143[m]> > I was just suggesting to drop the one_hot_encoding in the notebook, since categorical data is not supported
< jeffin143[m]> Got it ,
< jeffin143[m]> Yeah you can one hot encode a particular column , or multiple column
< jeffin143[m]> So you can first create a mapping of the string type and then one hot encode it
mindlifter has quit [Remote host closed the connection]
< jeffin143[m]> But like zoq suggest you can ditch the preprocess one hot encoding for string data type
< zoq> Not sure there are plans to support that?
mindlifter has joined #mlpack
< jeffin143[m]> I mean , we can't store it in mat ?? Right
< zoq> Good question, since arma::Mat can't store std::string, there are some thoughts around using xtensor wondering if that is one way to fix this.
< rcurtin[m]> jeffin143: in a binding you can use `PARAM_MATRIX_WITH_INFO`, which will give you a `DatasetInfo` along with the `arma::mat`
< rcurtin[m]> the automatic binding implementation for each language then defines what a user can actually pass in that type of situation
< jeffin143[m]> Ok
unclutter has joined #mlpack
unclutter has quit [Remote host closed the connection]
Cthulhux9 has joined #mlpack
Cthulhux9 has quit [Remote host closed the connection]
Thaodan_ has joined #mlpack
Thaodan_ has quit [Remote host closed the connection]
mindlifter has quit [Remote host closed the connection]
mindlifter has joined #mlpack
ImQ009 has quit [Quit: Leaving]
tnR has joined #mlpack
tnR has quit [Remote host closed the connection]
FinboySlick has joined #mlpack
FinboySlick has quit [Killed (Sigyn (Spam is off topic on freenode.))]
< rcurtin[m]> it looks like maybe CMake 3.20 broke the CI build? that's frustrating... can anyone reproduce it locally?
< heisenbuugGopiMT> What do I have to do to reproduce it? Can you elaborate a bit if that's possible...
< rcurtin[m]> I am not sure, but just looking at the output, it seems like just using CMake 3.20 at all to try and compile mlpack causes `make` to fail at the next step
< heisenbuugGopiMT> Okay, I am trying...
mindlifter has quit [Remote host closed the connection]
< rcurtin[m]> I just created a branch https://github.com/mlpack/mlpack/pull/2976, and we'll see what that does... I have no idea if it will help
< rcurtin[m]> I haven't actually reproduced the issue locally, but just digging through CMake files that seemed a little off
mindlifter has joined #mlpack
< heisenbuugGopiMT> I am building cmake 3.20 currently, once it's done, I will try and build mlpack...
< heisenbuugGopiMT> Also after making my latest commit, this test is failing...
< heisenbuugGopiMT> This seems a bit unrelated to my changes maybe...
< heisenbuugGopiMT> Regarding the parser I think `load_save_test` is the test that matters...
< rcurtin[m]> it's possible that test does use `data::Load()` somewhere, but at least personally I think it's probably best to not bother to investigate it until `load_save_test` is all passing
< heisenbuugGopiMT> okay, thank you...
tarzeau[m] has joined #mlpack
tarzeau[m] has quit [Remote host closed the connection]
< heisenbuugGopiMT> Hey @ryan:ratml.org I am getting an error when I am trying to build it...
< zoq> heisenbuugGopiMT: What you see is only a warning.
< zoq> you end up with a build file
< heisenbuugGopiMT> Oh okay, I don't know how I misinterpreted it...
< zoq> No issue with 3.20.3 here.
< heisenbuugGopiMT> So, why are the builds failing?
< heisenbuugGopiMT> I was able to build mlpack successfully using `cmake3.20.3`
< heisenbuugGopiMT> * I was able to build mlpack successfully using `cmake 3.20.3`
< rcurtin[m]> awesome, thanks for trying that heisenbuug (Gopi M Tatiraju)
< rcurtin[m]> I'm not sure why CI is failing then; CMake says it completes successfully, but then `make` fails as though no Makefile was generated
GopiMTatiraju[m] has joined #mlpack
< heisenbuugGopiMT> I ran the `make` command on my local system, it worked all right...
< GopiMTatiraju[m]> .
GopiMTatiraju[m] has left #mlpack []
mindlifter has quit [Remote host closed the connection]
mindlifter has joined #mlpack