cake93

This week, the PyMCon 2020 is happening: All around PyMC3, one of the leading libraries for MCMC/Bayesian statistics/probabilistic programming, this online conference is split into an asynchronous and a synchronous part.

The talks of the asynchronous part are publicly accessible on the forum: https://discourse.pymc.io/c/pymcon/2020talks/15

Tickets for the synchronous part with Keynote sessions and live Q&A are still available: https://pymc-devs.github.io/pymcon/

Yes.

These are images taken by confocal microscopy, where you focus a laser scanning microscope on a plane and only see that. Then you move the focus plane up by a bit and take the next picture.

Super-resolution microscopy such as STORM, STED or PALM on the other hand will give you coordinates.

[follow up]

For Light Sheet 3D video of developing fly embryo take a look here: http://www.digital-embryo.org/ They have movies and also downloads of the raw data and Matlab pipline for analysis.

Notice that this was published in 2012. Fresh data doesn't get published before the analysis is done and the paper written..

Here, follow the instructions to download and install the HDF5 plugin for Fiji:

http://lmb.informatik.uni-freiburg.de/resources/opensource/i...

From the same site you can download "pollen.h5". Then follow the steps to load the dataset in "hyperstack (multichannel)" mode. This will open a window where each frame in the movie is a z-slice of the 3D image.

Then go to Plugins/Volume Viewer (scroll down) and switch the mode (top left) to "Volume". (this is what you should get: http://i.imgur.com/0QoGa1t.png)

The same people also published lots of 3D data of the zebrafish embry. Have a look at:

http://vibez.informatik.uni-freiburg.de/

Look for "e098.h5" to download it.

If you have never worked with the HDF5 format before, you can use HDF5 View (https://support.hdfgroup.org/products/java/hdfview/) to look into the files. As you can see here (http://i.imgur.com/nnTiyQh.png) the data in pollen.h5 basically is a uint8-array of 193x199x419.

There are libraries for many languages to read HDF5. Eg. "h5py" for Python.

Manipulation with the mouse if you're lucky. Often this would involve Matlab, or some specialized software. But in many cases we don't even have this and people use Fiji to scroll through the z-dimension with a slider.

For presentations people often render a movie with Matlab, investing hours to get it right. With AR, you could take a movie by filming with a virtual camera in your hand.

Augmented Reality would add the most value. Some examples:

+ intuitive exploration of the data (imagine learning about a plant by scrolling through cross-sections)

+ intuitive manipulation of the perspective (the mouse 3D rotation thing is really tricky)

+ collaborative viewing

+ annotation of objects (eg. tracing a filament through 3D by following it with the finger)

+ avoid occlusion by just zooming and moving in

+ be able to point at things in a 3D image

It would bring much more natural ways to interact with the data. Essentially scaling up your molecular structure 10^6 to 10^7-fold so you can explore it as you'd explore a sculpture.

AR/VR visualization of 3D and 4D microscopy data (multicolor 3D video, as well as 3D point clouds over time) for biological research.

Look for "lattice light sheet microscopy" or "superresolution microscopy" such as (3D)-STORM or STED.

These techniques are adopted at a high pace. Groups spent $ 500,000 and often more on the hardware. They can produce terabytes of data within days, but we hardly have any tools to view and interact with it. (And the people are overwhelmed with the analysis.)

Imagine a holographic video of living cell (potentially in near real time) where you can zoom by grabbing the hologram.

Of course! This is cool in so many different ways.

I can imagine a large touch bar with stylus input (!) would actually be useful.

Also a trackpad at the top of the keyboard would be perfect for working on trains and planes!

Easy solution for problems 2+3: tiny USB sticks. For example SanDisk Cruzer Fit 32GB ($ 9) on amazon.

They have a tiny loop, but I prefer to always keep it plugged in to my laptop.

Always works.

Except with smartphones.

How would the calculations play out for near-earth MACHOs and radio frequencies instead of visible light? I can imagine that the construction of a suitable antenna is more feasible than for a visible-light telescope?

Correct me if I'm wrong, but from a thermodynamic point of view the major problem is the entropy. You can't just convert heat into any other form of energy (light) without also caring of the entropy. In the case of solving some salt in water, the entropy is going into the chaos of the solved ions (versus the low-entropy crystal). With all the other methods you need to get rid of the heat. With peltier coolers I could imaging "watercooling" the hot side.

On the other hand sweat (evaporation) is already a pretty good cooler and covering the skin with something else stops that effect. Additionally you enrich sweat inbetween which does not feel well. Completely absorbing the sweat also kills the cooling effect..

So maybe, as mchannon suggested, all you have to do is design a nice, flexible and good-looking sweatband with an integrated fan ;)

Biotechnologist here....As I see it there are three options: 1. Uptake of the thermic energy by an endothermic chemical reaction (like http://en.wikipedia.org/wiki/Instant_cold_pack) 2. Passive removal of the heat with a heat conducting material (you have to get rid of the heat somewhere else) 3. Active removal of the heat (see http://en.wikipedia.org/wiki/Cooling_vest)

I think thermoelectric cooling (http://en.wikipedia.org/wiki/Thermoelectric_cooling) would be suitable for your application. It is also compatible with Arduino technology I think.

The third party (reseller) who bought the seed from Monsanto in the first place most probably breached the contract by producing and selling new seeds! This third party is probably the one who is to be sued, but patent law allows the farmer to be sued as well..