I have a particle system, a material and I have 10 different textures to choose from. Is there a way to randomize the texture every 5 seconds for my emitter?
Use the Loop animation 'Looped' pulse to trigger the random value every time it completes a loop. Floor the random value so it outputs a whole number and then put the output from that into an Option Picker to choose the Texture.
patch example
Related
I'm trying to train an deep neural network model, the output dimensions of each iteration in one epoch is like [64,1600,8] (64 is the batch size). But in the last iteration of first epoch, this output changed to [54,1600,8] and faced with dimension error. Why in the last iteration batch size had changed??
Additionally, if I change the batch size to 32 the last iteration's output is [22,1600,8].
I think that the output of the last iteration must be same as the other iteration.
The last iteration batch size changed because you did not have enough data to completely fill the batch. If you have a batch size of 10, for example, and you have 101 entries total in your data, then you will have 10 batches of 10 and 1 batch of 1.
The solution is to either drop the batch if it is not the correct size, or to adapt your model so that it will detect the size of the batch and change accordingly, instead of having the batch size hard-coded in to your model parameters.
Seeing that you are using pytorch, I'll add to the answer by Richard by saying that pytorch DataLoaders have the functionality built-in to drop the last (incomplete) batch. Checking the documentation, you can specify drop_last=True while instantiating the DataLoader.
I am writing a Tetris program with PyGame, and came across a funny problem.
Before I ask the question, here is the pseudo-code:
while True:
# In this part, the human controls the block to go left, right, or speed down
if a key is pressed and the block isnt touching the floor:
if the key is K-left:
move piece left one step
if the key is K-right:
move piece right one step
if the key is K-down:
move piece down one step
# This part of the code makes the piece fall by itself
if the block isnt touching the floor:
move block down one step
# This part makes the while loop wait 0.4 seconds so that the block does not move
# down so quickly
wait 0.4 seconds
The problem is that, because of the "wait 0.4 seconds" part of the code, the part that the human controls can only move every 0.4 seconds. I would like it so that the block moves as fast as the human can press the key, while at the same time, the block dropping every 0.4 seconds. How could I arrange the code so that it will do that? Thanks!
The main problem I see here is that you are limiting your framerate using a wait of 0.4 seconds.
You should not limit framerate, but instead, you should limit how fast your block falls.
If I remember well, there was a formula you could use to do just that. It was based on the amout of time elapsed since the last frame. It looked like:
fraction of a second elapsed since last frame * distance you want your block to move in a second
This way, you can keep your mainloop intact, and the move processing will happen at every frame.
You could also do...
...
# This part of the code makes the piece fall by itself
if the block isn't touching the floor and
the block hasn't automatically moved in the last 0.4 seconds:
move block down one step
...
Just realize you'll be doing a lot of polling if the user hasn't struck any keys.
You may try asking gamedev.stackexchange.com instead. Check the site for Game Loops, and check out other example pygame projects to see how they're doing it. Having a good game loop is essential and will take care of things for you such as user inputs and a consistent frame rate.
Edit: https://gamedev.stackexchange.com/questions/651/tips-for-writing-the-main-game-loop
When doing games you should always try to do something like this:
while not finished:
events = get_events() # get the user input
# update the world based on the time that elapsed and the events
world.update(events, dt)
word.draw() # render the world
sleep(1/30s) # go to next frame
The sleep time should be variable so it takes into consideration the amount of time spend drawing and calculating the world updates.
The world update method would look something like this:
def update(self, events, dt):
self.move(events) # interpret user action
self.elapsed += dt
if self.elapsed > ADVANCE_TIME:
self.piece.advance()
self.elapsed = 0
The other way of implementing this (so you dont redraw too much) is to have events fired when the user orders a piece to be moved or when ADVANCE_TIME time passes. In each event handler you would then update the world and redraw.
This is assuming you want the pieces to move one step at a time and not continuous. In any case, the change for continuous movement is pretty trivial.
I've been playing with SysTick for a couple of days and i cannot reach nanoseconds delay. Is it possible with Systick to reach such small values or i have to use timers and interrupts? The LEDs though won't work lower than 350ns delay. Here is an image from my usb oscilloscope:
In general I want to make a project (i am just experimenting with LEDs and SysTick above) which will be like this:
where Δt = 250ns (the other parameters will be determined somehow). The question is, can I make these pulses by using SysTick?
STM32F407VG have 24-bit SysTick timer and its maximum clock speed is 168MHz (Core Clock speed). That means, even if you set your SysTick reload register to:
0x000001 (1 cycle)
You can only have 5.95ns period.
I found this in section 6.2 Clocks of the RM0368 reference manual:
The RCC feeds the external clock of the Cortex System Timer (SysTick) with the AHB clock (HCLK) divided by 8. The SysTick can work either with this clock or with the Cortex clock (HCLK), configurable in the SysTick control and status register.
So maybe the maximum tick rate is limited by the clock divisions. Check Figure 12. Clock tree to see which clock config you should use to get maximum speed.
I'm having trouble achieving frame rate independent motion in AS3 at 60fps. Every frame I measure the time since the previous frame, add it to an accumulator, and if the accumulator is greater than my target time, which is 16.666ms (60fps), a frame is simulated.
The problem is that the AS3 getTimer() only returns a time in milliseconds.
The delta times I get are often 16ms for the first frame, 16ms for the second, then 18ms for the third, and this pattern repeats. This averages out to 16.666. But in the first frame it is lower than the target time (16 < 16.666), so no frame is simulated. In the second frame the accumulator is higher than the target time, but slightly less than double it, so one frame is simulated. For the third frame 18ms pushes the accumulator over double the target time, so two frames are simulated.
So I'm getting this very jerky motion where no frames are rendered, then one, then two, then none, then one, then two, and this continues.
How would I get around this?
Wow... I thought I was only one who found that out.
Yes, the timer class in AS3 is not accurate. It will only trigger every ~16ms which causes MAJOR issues at times.
If want to see 2048 actions in 1000ms: FlashText Markup Language
(To test this, you'll need a method which takes 1ms to execute - just for even stats)
Notice the difference:
CORRECT: 1000ms | timer=0 == 1000 actions
AS3: 1000ms | timer=0 == 62.5 actions
I was able to write a class that works like this:
CORRECT: 1000ms | timer=0 == 1000 actions
RESHAPE TIMER: 1000ms | timer=0 == 1024 actions
NOTE:
it DOES NOT fire an action every ms
it DOES catch up to actions between the 16ms interval
an ACTION (the way I use it) is a method call, each method call can have its own body
The methodology I used to create this was catch-up... basically the first timer event will fire at 16ms... we know we have a full 16ms worth of code time to fire our own actions before the timer class fires again - thats where you inject sub-actions...
The highest I was able to produce was 2048 actions in 1000ms... more than 2 actions per ms.
Now... back to your problem
There is NO WAY to trigger a 0ms timer event. Based on my solution, if you want to by-pass the first 16ms of lag before the timer fires... you can dispatch an event which will fire within 2ms depending on the current system processes.
Possible Solution
If you take my approach for throwing your own actions within the 16ms, then you can build your own timer class. Use events for times under 16ms, when fired... fire 15 more - lol. Basically, you can create your own deltas between the 16ms.
I have a program running in real-time, with variable framerate, e.g. can be 15 fps, can be 60fps. I want an event to happen, on average, once every 5 seconds. Each frame, I want to call a function which takes the time since last frame as input, and returns True on average once every 5 seconds of elapsed-time given it's called. I figure something to do with Poisson distribution.. how would I do this?
It really depends what distribution you want to use, all you specified was the mean. I would, like you said, expect that a Poisson distribution would suit your needs nicely but you also put "uniform random variable" in the title which is a different distribution, anyway let's just go with the former.
So if a Poisson distribution is what you want, you can generate samples pretty easily using the cumulative density function. Just follow the pseudocode here: Generating Poisson RVs, with 5 seconds being your value for lambda. Let's call this function Poisson_RN().
The algorithm at this point is pretty simple.
global float next_time = current_time()
boolean function foo()
if (next_time < current_time())
next_time = current_time() + Poisson_RN();
return true;
return false;
A random variable which generates true/false outcomes in fixed proportions with independent trials is called a Geometric random variable. In any time frame, generate true with probability 1/(5*fps) and in the long run you will get an average of one true per 5 seconds.