Not giving the correct average while calculating the Mean Absolute Percentage Error

I am analysing a forecast and a station data. During the night, I am getting negative values on the station, and zero values on the forecast data. And there are some days I am missing some data.

Keeping that in mind, I am trying to calculate the Mean Absolute Percentage Error (MAPE) in DAX within Power BI to compare my forecast values to my station data. If there is a value equal or less than a 0, or there is not any value, then do nothing. However, I am not getting the correct average for my MAPE. Here's how I am currently calculating it:

MAPE =
VAR Actual =
    AVERAGE ( STATION_TABLE[GHI] )
VAR Forecast =
    AVERAGE ( FORECAST_TABLE[GHI] )
VAR AbsoluteError =
    ABS ( Actual - Forecast )
RETURN
    IF (
        OR ( Actual <= 0, Forecast <= 0 ),
        BLANK (),
        DIVIDE ( AbsoluteError, Actual )
    )

I am then averaging the MAPE for all of my data points using the following measure:

MAPE_Average = AVERAGE([MAPE])

I made a table with the date, hour, station data, the forecast data and the MAPE result to compare the result. The MAPE is correct until it calculate the average.

However, when I compare this value to the MAPE total calculated using other tools (such as Excel), I am getting a different average. The average on the table is the value MAPE result of the TOTAL AVE values at the end of the table. So, I am not getting the average of the whole column. And the measurement, which I did in apart, is giving another result which is not the same in Excel and neither the table.

Can anyone help me identify what I might be doing wrong, or suggest an alternative approach for calculating the MAPE in DAX?

Here I will add one day as example:

If we check the total average, it should be 13.52% a not 9.45% which is the MAPE between 226.95 and 205.50.

I found the answer but it can't be applied in all cases, I just create a new measure that call the measure I posted before:

Measure_2 = 
SUMX(
    VALUES('Fact Table_TIME'[Hour]),
    IF(
       [MAPE] > 0, [MAPE]
    )
)/12

The 12 is the counts of values in the column [MAPE]