Null sign Puns

I am trying to create a value in a report that takes First Year Amount * (Service Length in Months / 12)

First, the "Service Length in Months" field in our Opportunities is not required. For one year (12 month) deals, it is often left blank.

So, I started to create a new column with this IF Function logic test...

IF(logical_test, value_if_true, value_if_false) <- TEST

IF(Opportunity.Service_Length_months__c:SUM="-", 12, Opportunity.Service_Length_months__c:SUM) <- MY TEST

The logic here is that if the Service Length is left blank (ie: null, or "-") it would default to 12. However, I am getting an error for...

"Field expression does not exist. Check spelling."

Thank you for your help.

Im learning nonparametric tests and am having some trouble understanding the paired samples sign test. it makes sense for the single sample case that we would expect half the values to be above the median and half below and hence have equal probability of getting a + or - sign on the difference between the mean and each value. I cant seem to wrap my head around how this also applies to the differences of two samples. Im having trouble seeing how two random variables with the same median will have on an, equal number of positive differences and negative differences (a median of zero) . Can somebody elaborate and help provide intuition as to why this must be so

Any ideas what it could be? The null wisp is white with a black fire

Hi everyone.
I am implementing the PGSGP Google Play Games Services Plugin for Godot.
I get to do the singin correctly but when i print account_id this is null, the function does not retrieve the accountid.

This is what i have coded following PGSGP:

func _on_sign_in_success(account_id : String):     
print("_on_sign_in_success")     
GLOBAL.player_google_id = account_id     
print("account_id: " + account_id) 

This is the print:

0 param: --remote-debug 
1 param: localhost:6007 
2 param: --xr_mode_regular 
3 param: --use_depth_32 
4 param: --use_immersive 
Installing to device (please wait...): Samsung SM-A510F 
--- Device API >= 21; debugging over USB --- 
--- DEVICE API >= 21; DEBUGGING OVER USB --- 
Reverse result: 0 
--- Debugging process started 
--- Godot Engine v3.2.3.stable.official - https://godotengine.org 
OpenGL ES 3.0 Renderer: Mali-T720  

_on_sign_in_success 
account_id:  

what could be happening? I've tried everything ...

I am currently performing a classification analysis, which yields a few thousand values of classification accuracy stemming from real data for each of my participants' data (which will be called 'accuracy maps' from here on). Imagine I have 20 participants.

I further computed permutation samples as follows.

For each permutation 'run', I 'flip a coin' for every participant. If the coin should come heads, I leave the 'accuracy map' of a given participant the way it is. If it comes tails, I subtract the whole 'accuracy map' from 1. This gives me 20 'null' (as in no effect) accuracy maps. I then take the average map of these twenty, and store the biggest single accuracy-value, included in this average map somewhere.

I do the whole procedure 10000 times, which gives me 10000 values of classification accuracy. Crucially, these 10000 values are 'maxima under the null', meaning the maximum classification accuracy that was observed under the null.

I finally generate a distribution from these 10000 accuracies (picture: https://imgur.com/a/OLz2o).

The idea is to test each value of classification accuracy of the 'average real data' map vs. the distribution of maxima under the null. E.g., I am looking for a statistical test that answers, for each real value of average classification accuracy, 'is this value significantly higher than my null-distribution maxima?

It is easy for me to imagine this graphically. I would just put each 'real' accuracy into the distribution (see picture) and check whether its value corresponds to a value equal or higher than 95 or 97.5% of the distribution.

However, I am stuck at knowing which statistical test does this, as usually two distributions are tested versus each other, which is not the case for my setting.

I hope this was not too dubious. I appreciate any help.

Ran into some issues over the weekend with vcenter 5.5 Standard. I was able to get the fst client working and accepting authentication again but rhe web client is returning the above error when trying to log in as any user.

Is there any way to regenerate the sso token to re-establish the ability to log in with the web client, or is the only option a complete uninstall/reinstall of vcenter? If so, anything I can backup to preserve configs?

Noob here : I've been toying with the game for a few weeks, and I'm now getting bored of High Sec. So, I got my backpack and stuff, lots of Quafe, found a little station to unpack as near from Null as possible and now roam here. I read the sov maps to understand who's who here.

I spent some time roaming, learning the map, watching local, d-scan and such. I now wish to contact the (very few) active corps there and introduce myself to them. I'm not looking for blue status or whatever (even if I won't go after anyone whose corp hasn't gome after me before). But I just don't know how to introduce oneself to the real PVP corps nearby. How should I do that ?