This post almost answered this question for me, but I have a specific need and didn\'t find what I sought there. This lies right outside my experience; couldn\'t quite wrap
I may not have the best naming conventions, but here goes.
I calculate the intervals between the array of dates and the given date. I then do a sort, to find the "smallest" difference.
$dates = array
(
'0'=> "2013-02-18 05:14:54",
'1'=> "2013-02-12 01:44:03",
'2'=> "2013-02-05 16:25:07",
'3'=> "2013-01-29 02:00:15",
'4'=> "2013-01-27 18:33:45"
);
function find_closest($array, $date)
{
//$count = 0;
foreach($array as $day)
{
//$interval[$count] = abs(strtotime($date) - strtotime($day));
$interval[] = abs(strtotime($date) - strtotime($day));
//$count++;
}
asort($interval);
$closest = key($interval);
echo $array[$closest];
}
find_closest($dates, "2013-02-18 05:14:55");
If I understand your question perfectly then this will solve your problem.
<?php
$dates = array
(
'0' => "2013-02-18 05:14:54",
'1' => "2013-02-12 01:44:03",
'2' => "2013-02-05 16:25:07",
'3' => "2013-01-29 02:00:15",
'4' => "2013-01-27 18:33:45"
);
function closest($dates, $findate)
{
$newDates = array();
foreach($dates as $date)
{
$newDates[] = strtotime($date);
}
echo "<pre>";
print_r($newDates);
echo "</pre>";
sort($newDates);
foreach ($newDates as $a)
{
if ($a >= strtotime($findate))
return $a;
}
return end($newDates);
}
$values = closest($dates, date('2013-02-04 14:11:16'));
echo date('Y-m-d h:i:s', $values);
?>
Just try this:
$date = array(
[0]=> "2013-02-18 05:14:54"
[1]=> "2013-02-12 01:44:03"
[2]=> "2013-02-05 16:25:07"
[3]=> "2013-01-29 02:00:15"
[4]=> "2013-01-27 18:33:45");
$baseDate = date_create('2009-10-11');
$count = count($date);
for($loop=0;$count>$loop;$loop++) {
$datetime = date_create($date[$loop]);
$interval = date_diff($baseDate, $datetime);
$newDate[$interval->format('%s')] = $date[$loop];
}
ksort($newDate);
foreach($newDate as $key=>$value) {
echo $value;
break;
}
Your first element will the the closest match date.
Note: Please test it before you use.
Suppose your array is bigger and that you have dates over the period 2009-10-01
to 2019-10-01
. Lets now compare two approach: a. looping-array approach
vs b. sorting-indexing-array approach
.
<?php
$period = new DatePeriod(
new DateTime('2009-10-01 00:00:00'),
new DateInterval('P3D'),
new DateTime('2019-10-01 00:00:00')
);
foreach($period as $date){
$dates[] = $date->format('Y-m-d');
};
$today = '2019-08-18 13:00:15';
function custom_sort($array){
sort($array);
return $array;
}
function nearest_date_key($array, $today){
//push today to the array, sort the array,
//find the nearest day value of the sorted array and return key
array_push($array, $today);
$sorted_dates = custom_sort($array);
$find_today_key = array_search($today, $sorted_dates);
$nearest_date = array_slice($sorted_dates, $find_today_key + 1, 1);
return array_search($nearest_date[0], $array);
}
function find_closest($array, $today)
{
//$count = 0;
foreach($array as $day)
{
//$interval[$count] = abs(strtotime($date) - strtotime($day));
$interval[] = abs(strtotime($today) - strtotime($day));
//$count++;
}
asort($interval);
$closest = key($interval);
return $closest;
}
$start = microtime(true);
$result_a = nearest_date_key($dates, $today);
$time_elapsed_secs_a = microtime(true) - $start;
$start = microtime(true);
$result_b = find_closest($dates, $today);
$time_elapsed_secs_b = microtime(true) - $start;
?>
Printing the results gives (http://phptester.net/)
result time_elapsed
loop approach (a) 1203 0.00630
sorting index approach (b) 1203 0.00062
Which is a huge time elapsed gain. We divided by ten the waiting time