PHP: Remove a row from a CSV if ID occurs multiple times and a column has a specific value

Question

I have data as below in a CSV file. The green lines are what I'd like to keep.

Basically if someone has a single line, I want to keep it.

If they have multiple lines, I want to remove any where the 3rd column is A - Fully Fit.

After running through the whole file, I then want to save it over the original.

I tried writing the code but not sure if it's a Friday but the logic is escaping me at the moment.

There could be just a single line or there could be dozens. So if an ID has 1000 rows, I would just want to delete the rows for that ID where the 4th column is "A - Fully Fit"

UPDATE

This code works but not sure if it's the most optimal

// DECLARE ARRAYS
$a = Array();
$b = Array();
$c = Array();

// LOOP THROUGH FILE AND COLLECT DUPLICATES
if (($handle = fopen($filename, "r")) !== FALSE) {      

    while (($line = fgetcsv($handle, 4096, ",")) !== FALSE) {
        $a[$line[1]][] = 'SHAKKA';
    }

    foreach ($a as $k=>$v) {
        if (count($v)>1) {
            $b[] = $k;
        }
    }
}

// IF A DUPLICATE THEN REMOVE ROWS THAT ARE 'A - Fully Fit'
if (($handle = fopen($filename, "r")) !== FALSE) {      
    while (($golly = fgetcsv($handle, 4096, ",")) !== FALSE) {
        if (in_array($golly[1],$b) && $golly[3] == 'A - Fully Fit')  {
        }
        else {
            $c[] = $golly;
        }
    }
}
fclose($handle);

// WRITE THE FILE
$fp = fopen($filename, 'wa+');
foreach ($c as $fields) {
    fputcsv($fp, $fields);
}
fclose($fp);

Answer 1

<?php


if (($handle = fopen($filename, "r")) !== FALSE) {      
    $new_rows = [];
    $hash_ids = [];

    while (($line = fgetcsv($handle, 4096, ",")) !== FALSE) {
        if($line[3] === 'A - Fully Fit'){
            if(isset($hash_ids[$line[1]])) continue;
            $hash_ids[$line[1]] = true;
        }
        $new_rows[] = $line;
    }

    fclose($handle);

    // WRITE TO THE FILE
    $fp = fopen($filename, 'wa+');
    foreach ($new_rows as $current_row) {
        fputcsv($fp,$current_row);
    }
    fclose($fp);
}

In the above code, we maintain a hash(associative array) for each ID. If we have already come across with an ID with A - Fully Fit, then we skip the row, else we add it to the list.