The Indian radar remains obscure, confirmed by other source.

https://community.windy.com/topic/8191/your-animations/35130

@Suty why are you forcing it on us? Many of us don't want it
I don't want to have to "rewind" the loop manually and I don't want future cast.

If it stays I'll be forced to find an alternative as that will 100% block my use of this app.

x2 agreed

If it stays I'll be forced to find an alternative as that will 100% block my use of this app.

I don't mind it staying in the radar animation if it works properly, and does not mess up the end of the animation.

The prediction function appears to be working much better today, compared to how it has been working prior.

Others can test to confirm.

Note that the flickering signal at the bottom of this animation is the Longreach radar operating intermittently.

Am I mistaken, or is the prediction algorithm working better in the past few hours?

It seems to be doing a much better job of predicting individual rain feature vectors.

Before it was sending all the features in the same direction, but in this animation sequence I just made it is projecting all the rain features moving in different directions, consistent with their prior logged movement and speed vector.

If so, that's working properly (or at least it's a lot better).

Before the next release, just click on the white part of the timeline. Did you look at the @Suty’s post here above ?

Is it also possible to remove the forward movement prediction from radar data animations?

It makes a bit of a mess of them at present.

Yes, thank you, the animation is working properly again without jerking around, and is playing the data animation first once more. Which is fine and useful once again.

Though, the movement projection remains the same.

That's ok, I won't use the future movement option until a better movement prediction algorithm is developed (if possible).

Thanks, an example of what's occurring tonight.

@Sinewave1 We are still getting the feedback and we will try to do our best to optimize the behavior.

They cross each other at about 90 degrees, and at a similar movement speed.

So the prediction algorithm just spits the difference between them, and predicts they'll all move SSW!

No, none of these showers move SSW! They of course continue on their original headings. So this prediction is quite useless to me, it just averages the NET movement vector, rather than predicting movement and speeds of individual features, based on past movement and speed.

Plus it plays that 'prediction' first, instead of the logged data animation, which is actually predicatively very useful.

It's not a good feature.

Yup, not a fan. Don't like how it works, don't like what it does.

The movement predictions can also be wrong by up to 45 deg for storms and squalls.

Plus, it degrades the actual radar animation. Not worth having.

I'm talking about the regular "past" radar. It never used to wobble around like it is now, and can be seen in the above video.