Re: [Scilab-users] inserting data into a bigger matrix
Hi Denis, Thanks for sharing a neat and mind-blowing solution. It is much faster and with huge memory savings. Kind regards, Rafael From: users [mailto:users-boun...@lists.scilab.org] On Behalf Of CRETE Denis Sent: Wednesday, October 12, 2016 3:32 PM To: Users mailing list for Scilab <users@lists.scilab.org> Subject: Re: [Scilab-users] inserting data into a bigger matrix Hello, Assuming input data are stored in I = [0,0,0,1; 10, 0, 3, 1;15, 30, 0, 1;18, 0, 45, 0;36, 57, 28, 0]; there might be also a solution using the “sparse” function: Nb_Dat=size(I); Time= I(:,1); ij=[1+ones(Nb_Dat(2),1).*.Time,(1:Nb_Dat(2))'.*.ones(Time)] ; sp=sparse(ij,I); // Display full Matrix full(sp) and it may be more efficient (?) writing it on a single line: sp2=sparse([1+ones(size(I,'c'),1).*. I(:,1),(1: size(I,'c'))'.*.ones(I(:,1))],I); HTH. Denis De : users [mailto:users-boun...@lists.scilab.org] De la part de Philipp Mühlmann Envoyé : mercredi 12 octobre 2016 14:55 À : Users mailing list for Scilab Objet : Re: [Scilab-users] inserting data into a bigger matrix charming...exactly what I need. 2016-10-12 14:45 GMT+02:00 Rafael Guerra <jrafaelbgue...@hotmail.com<mailto:jrafaelbgue...@hotmail.com>>: Hi Philipp, Does the simple code here below meet the requirements? I = [0,0,0,1; 10, 0, 3, 1;15, 30, 0, 1;18, 0, 45, 0;36, 57, 28, 0]; ix1=I(1,1); ix2=I($,1); ix=ix1:ix2; M=zeros(length(ix),4); M(1:$,1) = ix'; M(I(:,1)+1,:)=I; PS: It should be easy to adapt it for a general time series consisting of floats and constant time-sampling. Regards, Rafael From: users [mailto:users-boun...@lists.scilab.org<mailto:users-boun...@lists.scilab.org>] On Behalf Of Philipp Mühlmann Sent: Wednesday, October 12, 2016 2:16 PM To: International users mailing list for Scilab. <users@lists.scilab.org<mailto:users@lists.scilab.org>> Subject: [Scilab-users] inserting data into a bigger matrix Dear Scialb users, how to insert time based data into a pre-defined Matrix without using a for-loop? The data points are not equally spaced in time. example: // assuming DATA includes 5 data Points // each dat Point consists of 4 variable // first variable = time //datafile could look like Time, Var1, Var2, Var3 0, 0, 0, 1 10, 0, 3, 1 15, 30, 0, 1 18, 0, 45, 0 36, 57, 28, 0 Assume that a timestep of 1 second is wanted. desired result should look like this: M = 0 0 0 1 1 0 0 0 2 0 0 0 3 0 0 0. ... 10 0 3 1 ... 15 30 0 1 ... 18 0 45 0 ... 36 57 28 0 so "M" is bigger than the original dataset. Again, I think I could do this using for-loops. This could be OK for small dataset, but maybe become slow for huge data sets (> 100'000 data points). Thanks, Philipp Thanks, Philipp -- In Kanada is' ka' na' da. Sonst wär' Kanada Jemanda. There we have the salad. ___ users mailing list users@lists.scilab.org<mailto:users@lists.scilab.org> http://lists.scilab.org/mailman/listinfo/users -- In Kanada is' ka' na' da. Sonst wär' Kanada Jemanda. There we have the salad. ___ users mailing list users@lists.scilab.org http://lists.scilab.org/mailman/listinfo/users
Re: [Scilab-users] inserting data into a bigger matrix
Hello, Assuming input data are stored in I = [0,0,0,1; 10, 0, 3, 1;15, 30, 0, 1;18, 0, 45, 0;36, 57, 28, 0]; there might be also a solution using the “sparse” function: Nb_Dat=size(I); Time= I(:,1); ij=[1+ones(Nb_Dat(2),1).*.Time,(1:Nb_Dat(2))'.*.ones(Time)] ; sp=sparse(ij,I); // Display full Matrix full(sp) and it may be more efficient (?) writing it on a single line: sp2=sparse([1+ones(size(I,'c'),1).*. I(:,1),(1: size(I,'c'))'.*.ones(I(:,1))],I); HTH. Denis De : users [mailto:users-boun...@lists.scilab.org] De la part de Philipp Mühlmann Envoyé : mercredi 12 octobre 2016 14:55 À : Users mailing list for Scilab Objet : Re: [Scilab-users] inserting data into a bigger matrix charming...exactly what I need. 2016-10-12 14:45 GMT+02:00 Rafael Guerra <jrafaelbgue...@hotmail.com<mailto:jrafaelbgue...@hotmail.com>>: Hi Philipp, Does the simple code here below meet the requirements? I = [0,0,0,1; 10, 0, 3, 1;15, 30, 0, 1;18, 0, 45, 0;36, 57, 28, 0]; ix1=I(1,1); ix2=I($,1); ix=ix1:ix2; M=zeros(length(ix),4); M(1:$,1) = ix'; M(I(:,1)+1,:)=I; PS: It should be easy to adapt it for a general time series consisting of floats and constant time-sampling. Regards, Rafael From: users [mailto:users-boun...@lists.scilab.org<mailto:users-boun...@lists.scilab.org>] On Behalf Of Philipp Mühlmann Sent: Wednesday, October 12, 2016 2:16 PM To: International users mailing list for Scilab. <users@lists.scilab.org<mailto:users@lists.scilab.org>> Subject: [Scilab-users] inserting data into a bigger matrix Dear Scialb users, how to insert time based data into a pre-defined Matrix without using a for-loop? The data points are not equally spaced in time. example: // assuming DATA includes 5 data Points // each dat Point consists of 4 variable // first variable = time //datafile could look like Time, Var1, Var2, Var3 0, 0, 0, 1 10, 0, 3, 1 15, 30, 0, 1 18, 0, 45, 0 36, 57, 28, 0 Assume that a timestep of 1 second is wanted. desired result should look like this: M = 0 0 0 1 1 0 0 0 2 0 0 0 3 0 0 0. ... 10 0 3 1 ... 15 30 0 1 ... 18 0 45 0 ... 36 57 28 0 so "M" is bigger than the original dataset. Again, I think I could do this using for-loops. This could be OK for small dataset, but maybe become slow for huge data sets (> 100'000 data points). Thanks, Philipp Thanks, Philipp -- In Kanada is' ka' na' da. Sonst wär' Kanada Jemanda. There we have the salad. ___ users mailing list users@lists.scilab.org<mailto:users@lists.scilab.org> http://lists.scilab.org/mailman/listinfo/users -- In Kanada is' ka' na' da. Sonst wär' Kanada Jemanda. There we have the salad. ___ users mailing list users@lists.scilab.org http://lists.scilab.org/mailman/listinfo/users
Re: [Scilab-users] inserting data into a bigger matrix
charming...exactly what I need. 2016-10-12 14:45 GMT+02:00 Rafael Guerra <jrafaelbgue...@hotmail.com>: > Hi Philipp, > > > > Does the simple code here below meet the requirements? > > > > I = [0,0,0,1; 10, 0, 3, 1;15, 30, 0, 1;18, 0, 45, 0;36, 57, 28, 0]; > > ix1=I(1,1); > > ix2=I($,1); > > ix=ix1:ix2; > > M=zeros(length(ix),4); > > M(1:$,1) = ix'; > > M(I(:,1)+1,:)=I; > > > > PS: > > It should be easy to adapt it for a general time series consisting of > floats and constant time-sampling. > > > > Regards, > > Rafael > > > > *From:* users [mailto:users-boun...@lists.scilab.org] *On Behalf Of *Philipp > Mühlmann > *Sent:* Wednesday, October 12, 2016 2:16 PM > *To:* International users mailing list for Scilab. <users@lists.scilab.org > > > *Subject:* [Scilab-users] inserting data into a bigger matrix > > > > Dear Scialb users, > > > > > > how to insert time based data into a pre-defined Matrix without using a > for-loop? > > > > The data points are not equally spaced in time. > > > > > > example: > > // assuming DATA includes 5 data Points > > // each dat Point consists of 4 variable > > // first variable = time > > > > //datafile could look like > > > > Time, Var1, Var2, Var3 > > 0, 0, 0, 1 > > 10, 0, 3, 1 > > 15, 30, 0, 1 > > 18, 0, 45, 0 > > 36, 57, 28, 0 > > > > Assume that a timestep of 1 second is wanted. > > > > desired result should look like this: > > > > M = > > 0 0 0 1 > > 1 0 0 0 > > 2 0 0 0 > > 3 0 0 0. > > ... > > 10 0 3 1 > > ... > > 15 30 0 1 > > ... > > 18 0 45 0 > > ... > > 36 57 28 0 > > > > so "M" is bigger than the original dataset. > > > > Again, I think I could do this using for-loops. > > This could be OK for small dataset, but maybe become slow for huge data > sets (> 100'000 data points). > > > > Thanks, > > Philipp > > > > > > > > > > Thanks, > > Philipp > > > > > > > > > > -- > > In Kanada is' ka' na' da. Sonst wär' Kanada Jemanda. > > > > There we have the salad. > > ___ > users mailing list > users@lists.scilab.org > http://lists.scilab.org/mailman/listinfo/users > > -- In Kanada is' ka' na' da. Sonst wär' Kanada Jemanda. There we have the salad. ___ users mailing list users@lists.scilab.org http://lists.scilab.org/mailman/listinfo/users
Re: [Scilab-users] inserting data into a bigger matrix
Hi Philipp, Does the simple code here below meet the requirements? I = [0,0,0,1; 10, 0, 3, 1;15, 30, 0, 1;18, 0, 45, 0;36, 57, 28, 0]; ix1=I(1,1); ix2=I($,1); ix=ix1:ix2; M=zeros(length(ix),4); M(1:$,1) = ix'; M(I(:,1)+1,:)=I; PS: It should be easy to adapt it for a general time series consisting of floats and constant time-sampling. Regards, Rafael From: users [mailto:users-boun...@lists.scilab.org] On Behalf Of Philipp Mühlmann Sent: Wednesday, October 12, 2016 2:16 PM To: International users mailing list for Scilab. <users@lists.scilab.org> Subject: [Scilab-users] inserting data into a bigger matrix Dear Scialb users, how to insert time based data into a pre-defined Matrix without using a for-loop? The data points are not equally spaced in time. example: // assuming DATA includes 5 data Points // each dat Point consists of 4 variable // first variable = time //datafile could look like Time, Var1, Var2, Var3 0, 0, 0, 1 10, 0, 3, 1 15, 30, 0, 1 18, 0, 45, 0 36, 57, 28, 0 Assume that a timestep of 1 second is wanted. desired result should look like this: M = 0 0 0 1 1 0 0 0 2 0 0 0 3 0 0 0. ... 10 0 3 1 ... 15 30 0 1 ... 18 0 45 0 ... 36 57 28 0 so "M" is bigger than the original dataset. Again, I think I could do this using for-loops. This could be OK for small dataset, but maybe become slow for huge data sets (> 100'000 data points). Thanks, Philipp Thanks, Philipp -- In Kanada is' ka' na' da. Sonst wär' Kanada Jemanda. There we have the salad. ___ users mailing list users@lists.scilab.org http://lists.scilab.org/mailman/listinfo/users
[Scilab-users] inserting data into a bigger matrix
Dear Scialb users, how to insert time based data into a pre-defined Matrix without using a for-loop? The data points are not equally spaced in time. example: // assuming DATA includes 5 data Points // each dat Point consists of 4 variable // first variable = time //datafile could look like Time, Var1, Var2, Var3 0, 0, 0, 1 10, 0, 3, 1 15, 30, 0, 1 18, 0, 45, 0 36, 57, 28, 0 Assume that a timestep of 1 second is wanted. desired result should look like this: M = 0 0 0 1 1 0 0 0 2 0 0 0 3 0 0 0. ... 10 0 3 1 ... 15 30 0 1 ... 18 0 45 0 ... 36 57 28 0 so "M" is bigger than the original dataset. Again, I think I could do this using for-loops. This could be OK for small dataset, but maybe become slow for huge data sets (> 100'000 data points). Thanks, Philipp Thanks, Philipp -- In Kanada is' ka' na' da. Sonst wär' Kanada Jemanda. There we have the salad. ___ users mailing list users@lists.scilab.org http://lists.scilab.org/mailman/listinfo/users
