## utm parameters – Google Analytics not showing accurate data of UTM-tags

My UTM-tags are being activated as they do show up in Google Analytics real-time traffics tab. I can exactly see my medium and source whenever I test them.

However, when I go into all campaigns the data is not being tracked properly according to the data we only got 678 euro’s in sales for a specific campaign that has been “tracked” with the UTM-tag. Google Analytics picture

However, when I look in my FB ads manager we clearly have more revenue and also way more orders with the ads that contain the tags.

This also didn’t happen only once but multiple times and the data is completely wrong.

I checked if it was maybe a mistake with the UTM-tag itself but it should have been set-up properly as we use a UTM-tag building tool.

So I was wondering, is there a way to fix this?

## api – How accurate are these Bitcoin price predictions, and do they have a JSON blob?

I’ve found this: https://longforecast.com/bitcoin-price-predictions-2017-2018-2019-btc-to-usd

Very interesting short-term numbers. I hope they are right.

However, for me to actually use these numbers in an automated fashion, for example to display “how much I will own in X months”, I need a more reliable data format than this HTML table. I can’t find a JSON blob/API.

And how accurate have they been historically? Is there a better source for Bitcoin price prediction?

(No, I’m not going to “waste my kids college fund” based on the advice, but it would be fun to have a little automated number showing how much I will have if things go as predicted in X months.)

## unity – How to align prefab to a accurate position with the best practice?

I’m making a plumber game, I have a pump prefab, and pipe section prefab, which allows the user to connect the pipe during game, the pipe needs to be connected seamlessly during game, which means pipe prefab needs to be aligned in pixel level, please check below

User needs to add more and more pipe to connect the with each other, pipe is fixed length prefab.I’ve got choices,

1.Is the pipe prefab instantiated during game play a good practice? Because it might have many pipe prefabs, if the pipe goes a long way to another position, than there will needs to be so many pipe prefabs, compared to have procedrually generated pipe mesh during runtime, then a long pipe will not have to be cut off to the same small pieces?Which is good way to adpot?

2.If I use the same pipe prefabs to fabricate the piping system, how to accurately connect each other in pixel level, I can make the whole piping system in blender, then cut the system into pieces as my prefabs, then import to Unity as my first choice, this can do pixel level connection, but it seems to have better choice solely solved pixel level connection in Unity?

3.If I do it by generating the mesh during game play, then it could be complex especially it is hard to make elbow pipe mesh during runtime, or could you please advise me of some known library or package can do the work?

## exchanges – Are the transactions displayed on “BitListen” really accurate?

Over the years, whenever there’s unusual Bitcoin activity, I look at: https://www.bitlisten.com/

But can these transactions really be accurate? I just saw numerous 100+ BTC ones, and just now a gigantic one of 946.92 BTC (!!).

According to my exchange rate calculator, that is:

``````BTC 946.92 = \$22,731,761.52 USD
``````

Almost twenty-three million US dollars… at once?

I realize that a lot of money is being traded right now, but even if somebody has that many Bitcoin to trade, would they really do it all in one go? Isn’t that very risk? Wouldn’t you split it up into at least hundreds of different transactions?

I don’t understand what kind of person or entity would make a single transaction of such a huge sum. Is this really accurate or faked somehow?

## Accurate numerical calculation of nested integrals

I’d like to accurately (numerically) compute nested integrals of the type $$I(a;k) =left( prod_{i=1}^k int_0^infty mathrm{d}x_i right) exp left( – frac{1}{4a} sum_{i=1}^k left( x_i – x_{i+1} right)^2 – frac{a}{2} sum_{i=1}^k (x_i + x_{i+1}) right)$$ where $$a > 0$$ is a parameter and $$x_{k+1} equiv x_1$$. For $$k = 1$$ and $$k = 2$$ the integrals can be done analytically, but as far as I’m aware no closed-form expression is known for $$k > 2$$. So far I’ve written the following naive NIntegrate command:

``````int(a_, k_) := NIntegrate(Exp(-1/(4 a) (Sum((x(i) - x(i + 1))^2, {i, 1, k - 1}) + (x(k) -
x(1))^2) - a/2 (Sum(x(i) + x(i + 1), {i, 1, k - 1}) + (x(k) + x(1)))), ##, WorkingPrecision -> 30, PrecisionGoal -> 5) & @@ Table({x(i), 0, (Infinity)}, {i, 1, k});
``````

This command, for $$k = 3$$, gives an error (a `NIntegrate::slwcon`) when $$a$$ is small enough, $$a lesssim 1/3$$. For $$k = 4$$ the error appears when $$a lesssim 4/5$$. But even for larger $$a$$ when the error message is absent, I’m not sure if the command is computing the integral correctly. I’ve tried increasing the WorkingPrecision (as suggested by the error), but I’m finding it doesn’t help. The other suggestions are not relevant to this problem, I believe (no singularities or heavily oscillations, and the integral is not zero).

## Should Canon 5D mk II autofocus be accurate enough for a f/1.2 lens?

According to this Blog post by LensRentals.com owner Roger Cicala, the 5DIII has less than 1/2 the standard deviation of the 5DII with regards to Phase detection Auto Focus. But to get that additional benefit, one must also use Canon lenses introduced since about 2010. All of the testing was done using the center focus point.

In general, using the center focus point should yield faster, more accurate focus with faster lenses such as the 50mm f/1.2L and f/1.4. The 5DII can be an exception to this rule. When shooting in AI Servo AF, if C.Fn III-07 is enabled, there are six small AF Assist Points clustered within the spot metering circle that are active in addition to the Center Focus Point. The location of these six AF Assist Points are not indicated in the viewfinder! Yet the camera may be using one of them pointed at something that has higher contrast than what is located behind the square in the middle of the viewfinder when it decides the focus distance.

Here’s the Map for the 5DII’s focus system.

The entire series will tell you just about everything you’ve ever wanted to know about phase detection AF accuracy and consistency.

Autofocus Reality Part 1: Center-Point, Single-Shot Accuracy

Autofocus Reality Part II: One vs. Two, Old vs. New

Autofocus Reality Part 3a: Canon Lenses

Autofocus Reality Part 3b: Canon Cameras

Autofocus Reality Part 4: Nikon Full Frame

## MacBook Pro Activity Monitor accurate?

Unfortunately the answer is most likely “yes.”

There is a reason, or at least a possible to likely reason(s).

This could easily be bound to a failing component (logic board, drive, I/O etc.) As if there are hardware problems those won’t show up in activity monitor. Ask someone who (very recently) knows.

1. Boot the Mac while holding the d to go into diagnostic mode. The Mac will do its thing and report any problems it can.
2. Boot the Mac into Recovery Mode by holding the r kees at boot.

Once in recovery mode go to Disk utility. Select the top drive (your boot drive) and click on First aid.

Both of those tests will find gross issues with your Mac. It is not a guarantee that it will find a problem with any hardware or soft errors/bad blacks on the SSD. But it is always with trying.

OK so we’ve tried diagnostic tools, how about a different user profile? Go into System Preferences, create a new admin user and log in as that user. Problem gone? The issue was something in your user profile. The fun part is finding what it was. And by fun, I mean no fun at all.

If nothing is found there. I would back up your whole drive (Time Machine or 3rd party utility) Create a flash drive macOS installer with your preferred OS version (there are tutorials all over the place on hot to do this). Boot from the installer, re-format your HD and install the OS fresh.

Once that is done boot your mac normally and before restoring any of your apps, docs, pics or settings see if the problem is still there.

If the problem still exists it is most likely hardware and time to either have it repaired or buy a new Mac.

## tensorflow – Is the final history object of sequential model.fit() calls accurate?

I have several seperate training datasets that when read as one crash the kernel in jupyter. So I did a workaround and read them separately in a sequence and call fit() on the same model-object.

To get accuracy metrics I am only grabbing the final history-object, but does this also represent all previous fit()-calls?

## product photography – Which camera profile to use with SpyderCheckr in Lightroom/Adobe Camera Raw for most accurate colors?

For most uses, the standard color profile for your camera (“Adobe Standard”) should be sufficient. However, you may prefer the look of one of the other profiles for specific images. According the the SpyderCheckr (Windows, Macintosh) help files:

It is recommended that the “Adobe Standard” option be selected in the camera profile popdown list in Lightroom and ACR. This is the default choice, and the only option available in most cases, unless canned or custom profiles have been imported by the user. It is possible to use any profile as the basis for SpyderCheckr calibration, but its important to remember what profile was in place when the calibration was created, and to use that same profile when processing images through the resulting camera calibration preset. So the simplest solution is to leave the DNG profile at its default.

For “most accurate colors”, you would need to create custom camera profiles by photographing a color calibration chart and processing it with appropriate software to create an ICC or DCP file. A different profile for each camera and lighting setup would be needed.

How often to reshoot the chart depends on your needs. At the beginning of each session and whenever the lights are changed (position, amount, type, etc) would be reasonable. You don’t have to create new profiles if they aren’t needed, but you won’t be able to if you don’t have photos of the chart.

Other software that may be of interest: