Month: October 2017

Colleagues, programmers. Could you direct me to some proper material to read about automatic trend detection in data?

For example, if you have a log of events – say, temperatures from 10,000 sensors. We need to identify which sensor suddenly began to rise rapidly.

The first thing that comes to mind is to find trends over a short period of time and analyze micro-trends for two or three periods, but this approach has plenty of downsides: starting with the fact that there could be fluctuations not related to growth, and secondly, some sensors might rarely show readings compared to the analysis period, which causes a problem in correctly choosing the time period for finding the average. Essentially, this approach only works with very high-density information on sensors. And here, the density fluctuates – thick at times, then sparse for different sensors. Well, okay, you can make dynamic groups and somehow tag sensors as “frequent” and “rare”. But all this complicates things, and I feel like my thought is heading the wrong way.

Essentially, it is necessary to construct first and second order derivatives over time and analyze their shapes. Another problem is that the number of sensors is generally unlimited – some may appear, others disappear. Generally, new ones should also be trending.

What to read?

(TIL) The third largest city in California is called California City. Almost no one lives there, but its area is more than half a million square kilometers. It’s a ghost town. There are roads, on a huge territory, streets and avenues, but from horizon to horizon, you won’t see a single house. In the 50s of the last century, this city was a project of a millionaire, but the project did not come to a successful conclusion. In the end, everything remained as it was. The video is interesting

Continuing the theme of space: it would be interesting to conduct a survey on “how many times have Americans landed on the Moon?”. For some unclear reason, the vast majority do not know the correct answer (not you, of course), and are even very surprised by it.

I highly recommend the movie “Spacewalk”. Just rewatched it with the kids. It well portrays that much in the Soviet Union relied on heroism, often thoughtless, when human life was valued less than the country’s image.

Checked Wikipedia. I don’t know how to comment on what I found. In May ’65 Leonov performed a spacewalk, in ’69 it was Khrunov and Yeliseev, and in ’77 Gretchko and Romanenko. That’s it. Five people over 12 years, three flights. During these 12 years, about ~30 Americans from ~15 different spacecrafts (including six moon landings) did the same. It seems we started losing the space race right after that event…

Almost like in Moscow, January 1990: a very long queue at McDonald’s in Annandale, VA. All because today is the only day when McDonald’s brought back the Szechuan sauce to their menu, which hasn’t appeared there since 1998. Tomorrow it will be too late. Yesterday it was still too late.

There is this animated series “Rick and Morty”. In one of the episodes, Rick admits that the sole purpose of his space travels is the Szechuan teriyaki sauce that McDonald’s sold in 1998 as part of a collaboration with Disney to celebrate the release of the movie “Mulan”. McDonald’s came up with a cool trick – they sent the series creator a special box of the sauce with a note saying they made it happen, and they found a parallel dimension where it’s always 1998, and here, the sauce is yours, and also a little bit goes to the fans online.

The story is not about McDonald’s, but about marketing, really. Following in the footsteps of Nike, Adidas, and Harley Davidson, McDonald’s has moved from being a manufacturing company to a marketing company, where the product is not as important as the idea or the legend behind it. Nike, for example, also generates hype: a notification pops up on your phone that a limited collection of sneakers is on sale nearby (and yes, they are expensive), and you can make an online reservation right from the app (but you have to pick them up in a crowded place on the designated day, and they might even run out). A very clever idea. Now, traces of this are appearing in McDonald’s too.

There is one idea that could be turned into a decent game for iPads/iPhones. It would gather a fan base from all around the world. I once tried to create something on this topic, but lacked the hands/time/energy/desire. Maybe someone else will take it on.

The idea is to recognize a pattern tapped by a finger on the screen, and after two or three repetitions, catch and reproduce it through the program, with enriched sound and a steady tempo. Essentially, this would allow for the creation of a smart metronome for singing or accompanying on another instrument. It’s very simple to implement, but nobody has done it yet.

But that’s not all. The second part is to try to recognize the pitch of the device owner’s voice at the time and calculate an auto-accompaniment for the corresponding pattern. This is more complicated, but the task doesn’t look extremely difficult. Yes, of course, unlike the previous idea, this one requires not just a sense of rhythm but also a voice and musical ear, though both are not that rare nowadays (although I personally lack a singing voice).

Theoretically, for both tasks, there are simple solutions, and also not fully clear to me solutions involving machine learning – one could try using patterns taken from existing compositions, or find something in a library, if it already exists there. In any case, it’s a great topic for research at some university.

It would perfectly suit pop music. The system should easily pick up on something like “three chords” and offer a fairly rich assortment of accompaniments for the given rhythm and melody.

Ivan Shapovalov, Maxim Antonov, what do you think?

I still haven’t written about “Zaryadye”. I visited. Why is everyone thrilled? Didn’t like it. It seems like a sloppy job in almost every detail. If it wasn’t for the 14 billion, it could perhaps be forgiven. I just looked at Lebedev’s project, who is very proud of the navigation and fonts there. The navigation is okay, but the execution… Note that in Lebedev’s project, the stands for signs were supposed to be made of solid oak, but in reality, they made them from plywood, which started warping the very next day. And those painted manholes on the grass. And those unevenly laid tiles… All the structures in the park remind me of an ordinary shopping center somewhere in Kolomna. Concrete with glass. They could have created something pseudo-historical – the Kremlin is nearby after all. The views from the park are cool, that’s undeniable. A few weeks ago, I wrote about some real or fake news about the park’s accessibility for disabled people, which immediately started a backlash. Well, here’s a photo of the stairs, for example. Right after the entrance. And what about wheelchair users?..

It’s well known that i18n stands for internationalization, l10n stands for localization, gr8 stands for great, and 1ce stands for once.

Here’s a riddle for you: What does this numeronym mean? It just drove past me. To avoid exposing the answers, let’s agree that comments should only be in Russian translation, so I know you’ve figured it out.

Some features of American apartments would do well to be adopted in Russia. It seems to me that they are based on correct and convenient ideas, and do not have many downsides.

Firstly, there is little ceiling lighting here. As a result, there’s more pleasant “lower light,” and from the street, the house looks neater, as chandeliers do not blaze.

Secondly, the windows here start from the floor and go up to the ceiling. I understand that in cold Moscow this is not an option, but besides cold Moscow there are places like Krasnodar. It’s very convenient and bright.

The windows themselves open upward, by sliding, not swinging out. Surprisingly, the mechanism works entirely on friction. The window does not fall under its own weight, but you apply just a little force and it closes. These windows have advantages: no additional space is needed inside the apartment for opening windows, there’s no problem with the wind slamming the windows shut, and the windows look neater from both inside and outside. Our windows are open almost fully for weeks. I can’t imagine that with swinging windows. Similarly, the door to the balcony – it slides to the side.

In addition, there are no window sills, which for me are only pluses (otherwise they would have been cluttered with stuff).

Thirdly, instead of cupboards there are niches in the walls, which completely solve the storage issue and save living space.

Fourthly, there are no wallpapers here. Wallpapers are quite difficult to keep in order for years. And the walls are well designed even without wallpapers.

I apologize, but even the structure of local toilets is more sensible than in Europe.

There are also plenty of downsides. I do not understand this idiocy with the shower fixed in the wall, for example. But why not adopt what is good?

— And what have you achieved in life?