3 ms·

Computational Physics projects

7 points by aroman_ro 8y ago

For some time I started to implement (in C++) some projects related to computational physics. I thought they might be interesting for some, so I'll try to expose them here. The GitHub repositories are here: https://github.com/aromanro?tab=repositories the projects being described on the blog: https://compphys.go.ro

gus_massa 8y ago
Clicky: https://compphys.go.ro https://compphys.go.ro It's much better to submit each post individually than the whole page. It generates better (and more discussion). Also, post with URL only, because "self" post have a penalty here. I recommend posting one article per week, so people would not get annoyed and complain. (It's possible to make more submissions, but I prefer a lower rate.) Try to select a day/hour when you can be around in case people ask questions, it is very appreciated. You can also post a comment saying that you are the author, but it is not necessary. A nice technical post may get 5-10 upvotes and a few comments, and a few hundred visits to the page. Sometimes you are lucky and get more, sometimes a good post just falls thru the cracks and is ignored... I recommend to select the posts that have a nice visual part that is easy to understand. After people is hooked, they will read the technical part. I recommend to post first https://compphys.go.ro/lattice-boltzmann/ https://compphys.go.ro/lattice-boltzmann/ and https://compphys.go.ro/electric-field-lines/ https://compphys.go.ro/electric-field-lines/ Try to post once a week one of the old articles, and the new articles when they are published. You can make some repost later, but only a few times (2 or 3?). PS: I'd like to reed a follow-up of https://compphys.go.ro/nuclear-magnetic-resonance-and-fourier-transform/ https://compphys.go.ro/nuclear-magnetic-resonance-and-fourie... with some low-pass/high-pass filtering and other frequency based effects. Also, for the images in the frequency domain sometimes it is useful to peek and exponent n and draw \hat{f}^n(wx,wy) instead of \hat{f}^(wx,wy). Pick a small n like 1/2 or less to enhance small values, otherwise the peak at 0 is too big. Something similar to https://en.wikipedia.org/wiki/Gamma_correction https://en.wikipedia.org/wiki/Gamma_correction
- aroman_ro 8y ago
  The NMR project already has some filtering in place, nothing fancy, just cutting out high and/or low frequencies.

7 points by aroman_ro 8y ago

gus_massa 8y ago

Clicky: https://compphys.go.ro https://compphys.go.ro It's much better to submit each post individually than the whole page. It generates better (and more discussion). Also, post with URL only, because "self" post have a penalty here. I recommend posting one article per week, so people would not get annoyed and complain. (It's possible to make more submissions, but I prefer a lower rate.) Try to select a day/hour when you can be around in case people ask questions, it is very appreciated. You can also post a comment saying that you are the author, but it is not necessary. A nice technical post may get 5-10 upvotes and a few comments, and a few hundred visits to the page. Sometimes you are lucky and get more, sometimes a good post just falls thru the cracks and is ignored... I recommend to select the posts that have a nice visual part that is easy to understand. After people is hooked, they will read the technical part. I recommend to post first https://compphys.go.ro/lattice-boltzmann/ https://compphys.go.ro/lattice-boltzmann/ and https://compphys.go.ro/electric-field-lines/ https://compphys.go.ro/electric-field-lines/ Try to post once a week one of the old articles, and the new articles when they are published. You can make some repost later, but only a few times (2 or 3?). PS: I'd like to reed a follow-up of https://compphys.go.ro/nuclear-magnetic-resonance-and-fourier-transform/ https://compphys.go.ro/nuclear-magnetic-resonance-and-fourie... with some low-pass/high-pass filtering and other frequency based effects. Also, for the images in the frequency domain sometimes it is useful to peek and exponent n and draw \hat{f}^n(wx,wy) instead of \hat{f}^(wx,wy). Pick a small n like 1/2 or less to enhance small values, otherwise the peak at 0 is too big. Something similar to https://en.wikipedia.org/wiki/Gamma_correction https://en.wikipedia.org/wiki/Gamma_correction

aroman_ro 8y ago
The NMR project already has some filtering in place, nothing fancy, just cutting out high and/or low frequencies.