Calculate distinct count total at a group level

Hi, 

I'm trying to create a measure to get the "Other" count for each ethnicity but having trouble calculating total distinct count group by Ethnicity. There is 1:M relationship between Person(Ethnicity) and Behavior(Incident ID). 5,928 is a distinct count without ethnicity group and 6,297 is the total (sum of #Incident for each ethnicity) I'm looking for to determine Other. Ex: #Incident for Asian = 103, Other = 6297-103 = 6194. 

Expected result for #Incident Other

TotalAll-Wrong = CALCULATE(DISTINCTCOUNT('Behavior'[DISCIPLINE_INCIDENT_ID]), ALL(Person[Ethnicity]))

I was able to get 6297 as a total with these formulas but not when the ethnicity filter context was applied. 

# Incident All 1 = 
SUMX(
    VALUES(Person[Ethnicity Cd]),
    CALCULATE(DISTINCTCOUNT('Behavior'[DISCIPLINE_INCIDENT_ID])))

# Incident All 2 = 
VAR TotalGrpbyEthn = 
SUMMARIZE(
    'Person',
    Person[Ethnicity],
    "EthnDistinct", DISTINCTCOUNT ('Behavior'[DISCIPLINE_INCIDENT_ID] )
)
RETURN SUMX(TotalGrpbyEthn, [EthnDistinct])

# Incident All 4 = 
VAR TotalGrpbyEthn = 
SUMMARIZE(
    'Person',
    Person[Ethnicity],
    "EthnDistinct", DISTINCTCOUNT ('Behavior'[DISCIPLINE_INCIDENT_ID] )
)
RETURN CALCULATE((SUMX(TotalGrpbyEthn, [EthnDistinct])), ALLEXCEPT(Person, Person[Ethnicity]))

Relationship

I need this to be dynamic to work across multiple other slicers from School and Person tables on the report page. Could you please point me in the right direction? 

pbix file with sample data