由于每个数组的长度未知，我需要重新排列以下多维数组以合并和填充索引。问题是，我不知道如果数组结构错误，将需要多少行或多少列。
我首先需要一个数组，其中包含字符串用户ID，研究ID，分离中的每个研究的问题，然后是日期。我不必担心这个数组，因为创建起来非常简单。
我遇到麻烦的地方是，每个User_Prediction需要一个数组，其中它们包含的长度与问题数量相同（这有意义吗？）。我最终要寻找的是创建一个CSV文件，该文件的标题由UserId，StudyId和所有问题的组合列表以及日期组成。因此，我可以采用下面的“所需输出”，并为每个数组fcsvput简单地说。如果我现在按照User_Predictions的原样使用，则在水平排列（例如CSV文件）时，它们不会与问题对齐。
这是一个示例数组：
[studies] => Array
(
[0] => Array
(
[Study_ID] => 563
[Questions] => Array
(
[1] => S563 Q1
[2] => S563 Q2
)
[User_Predictions] => Array
(
[0] => Array
(
[User_ID] => 24
[Answers] => Array
(
[0] => 66
[1] => Option 1
)
)
)
)
[1] => Array
(
[Study_ID] => 510
[Questions] => Array
(
[1] => S510 Q1
[2] => S510 Q2
[3] => S510 Q3
[4] => S510 Q4
[5] => S510 Q5
)
[User_Predictions] => Array
(
[0] => Array
(
[User_ID] => 76
[Answers] => Array
(
[0] => 1
[1] => 1
[2] => Negative
[3] => 10
[4] => 2,2
)
)
[1] => Array
(
[User_ID] => 58
[Answers] => Array
(
[0] => 25
[1] => 20
[2] => Positive
[3] => 5000
[4] => 1,2
)
)
)
)
)
这就是我要实现的目标：
Array
(
[0] => User_ID
[1] => Study_ID
[2] => S563 Q1
[3] => S563 Q2
[4] => S510 Q1
[5] => S510 Q2
[6] => S510 Q3
[7] => S510 Q4
[8] => S510 Q5
[9] => Date
)
Array
(
[0] => 24
[1] => 563
[2] => 66
[3] => Option 1
[4] =>
[5] =>
[6] =>
[7] =>
[8] =>
[9] => Date
)
Array
(
[0] => 76
[1] => 510
[2] =>
[3] =>
[4] => 1
[5] => 1
[6] => Negative
[7] => 10
[8] => 2,2
[9] => Date
)
Array
(
[0] => 58
[1] => 510
[2] =>
[3] =>
[4] => 25
[5] => 20
[6] => Positive
[7] => 5000
[8] => 1,2
[9] => Date
)
答案 0 :(得分：3)
注意 ：不确定date的来源。
首先，这是您的问题模式。
studies中的每个数组都有一个Study_ID和一组问题Questions和许多用户的用户谓词User_Predictions。User_ID，并回答上一步中的问题。index of question = index of answer + 1 您的问题：您希望每个用户预测都作为一个单独的数组，但如果该特定研究中存在问题，则应包括所有问题列表及其答案。
解决方案：
为了进行预测，通常会有一个Study_ID和一个对此用户做出响应的用户User_ID。
// This array will contain all indexes should be present in final array.
$arrIndex = ['Study_ID', 'User_ID']; // Assume these two would always be present, so added in advance.
// This will be temporary array holds data for each user prediction with relevant information i.e. it would not have data for questions which are not in this study.
$arrUserStudy = [];
foreach($arr['studies'] as $arrStudy){
foreach($arrStudy['Questions'] as $strQuestion){
// To add a new question to the index.
if(!in_array($strQuestion, $arrIndex)){
$arrIndex[] = $strQuestion;
}
}
foreach($arrStudy['User_Predictions'] as $userStudy){
$arrTemp = [
'Study_ID' => $arrStudy['Study_ID'],
'User_ID' => $userStudy['User_ID'],
];
foreach($userStudy['Answers'] as $intKey => $strAnswer){
$arrTemp[$arrStudy['Questions'][$intKey + 1]] = $strAnswer;
}
$arrUserStudy[] = $arrTemp;
}
}
输出：
print_r($arrIndex);
Array
(
[0] => Study_ID
[1] => User_ID
[2] => S563 Q1
[3] => S563 Q2
[4] => S510 Q1
[5] => S510 Q2
[6] => S510 Q3
[7] => S510 Q4
[8] => S510 Q5
)
print_r($arrUserStudy);
Array
(
[0] => Array
(
[Study_ID] => 563
[User_ID] => 24
[S563 Q1] => 66
[S563 Q2] => Option 1
)
[1] => Array
(
[Study_ID] => 510
[User_ID] => 76
[S510 Q1] => 1
[S510 Q2] => 1
[S510 Q3] => Negative
[S510 Q4] => 10
[S510 Q5] => 2,2
)
[2] => Array
(
[Study_ID] => 510
[User_ID] => 58
[S510 Q1] => 25
[S510 Q2] => 20
[S510 Q3] => Positive
[S510 Q4] => 5000
[S510 Q5] => 1,2
)
)
如果将此输出与您的输出进行比较，则只有以下差异：
key作为问题/ Study_Id / User_ID，但是在您的输出中，它是一个整数。我们可以轻松地包含缺失的问题，因为我们确实在$arrIndex中包含了所有问题，而相应的值将是null。
将字符串键更改为整数键：我更喜欢使用字符串键，因为即使它们没有顺序，您也可以使用其key_name轻松识别和访问它们。
最终转换为所需输出：
保留字符串键并添加缺少的字符（首选）：
foreach($arrUserStudy as $intKey => $userStudy){
foreach($arrIndex as $strIndex){
if(!array_key_exists($strIndex, $userStudy)){
$arrUserStudy[$intKey][$strIndex] = null;
}
}
}
print_r($arrUserStudy);
Output:
Array
(
[0] => Array
(
[Study_ID] => 563
[User_ID] => 24
[S563 Q1] => 66
[S563 Q2] => Option 1
[S510 Q1] =>
[S510 Q2] =>
[S510 Q3] =>
[S510 Q4] =>
[S510 Q5] =>
)
[1] => Array
(
[Study_ID] => 510
[User_ID] => 76
[S510 Q1] => 1
[S510 Q2] => 1
[S510 Q3] => Negative
[S510 Q4] => 10
[S510 Q5] => 2,2
[S563 Q1] =>
[S563 Q2] =>
)
[2] => Array
(
[Study_ID] => 510
[User_ID] => 58
[S510 Q1] => 25
[S510 Q2] => 20
[S510 Q3] => Positive
[S510 Q4] => 5000
[S510 Q5] => 1,2
[S563 Q1] =>
[S563 Q2] =>
)
)
注意：这里您看到密钥没有排序，但是您可以循环$arrIndex并可以通过$arrUserStudy[<loop_variable>][<index_name>]在任意位置访问密钥。
转换为整数键并添加缺失的字符：
$arrReindexed = [];
foreach($arrUserStudy as $intKey => $userStudy){
$arrTemp = [];
foreach($arrIndex as $intKey => $strIndex){
$arrTemp[$intKey] = array_key_exists($strIndex, $userStudy) ? $userStudy[$strIndex] : null;
}
$arrReindexed[] = $arrTemp;
}
print_r($arrReindexed);
Output:
Array
(
[0] => Array
(
[0] => 563
[1] => 24
[2] => 66
[3] => Option 1
[4] =>
[5] =>
[6] =>
[7] =>
[8] =>
)
[1] => Array
(
[0] => 510
[1] => 76
[2] =>
[3] =>
[4] => 1
[5] => 1
[6] => Negative
[7] => 10
[8] => 2,2
)
[2] => Array
(
[0] => 510
[1] => 58
[2] =>
[3] =>
[4] => 25
[5] => 20
[6] => Positive
[7] => 5000
[8] => 1,2
)
)
答案 1 :(得分：1)
这是获得所需结果的相对简单的解决方案。
本文的末尾有一些说明。
这是输入数据的数组：
$studiesData = Array
(
"0" => Array
(
"Study_ID" => "563",
"Questions" => Array
(
"1" => "S563 Q1",
"2" => "S563 Q2"
),
"User_Predictions" => Array
(
"0" => Array
(
"User_ID" => "24",
"Answers" => Array
(
"0" => "66",
"1" => "Option 1"
)
)
)
),
"1" => Array
(
"Study_ID" => "510",
"Questions" => Array
(
"1" => "S510 Q1",
"2" => "S510 Q2",
"3" => "S510 Q3",
"4" => "S510 Q4",
"5" => "S510 Q5"
),
"User_Predictions" => Array
(
"0" => Array
(
"User_ID" => "76",
"Answers" => Array
(
"0" => "1",
"1" => "1",
"2" => "Negative",
"3" => "10",
"4" => "2,2"
)
),
"1" => Array
(
"User_ID" => "58",
"Answers" => Array
(
"0" => "25",
"1" => "20",
"2" => "Positive",
"3" => "5000",
"4" => "1,2"
)
)
)
)
);
修复数组的索引以使用有序键将它们全部获取：
foreach ($studiesData as $k => &$studyData)
{
$studyData['Questions'] = array_values($studyData['Questions']);
foreach ($studyData['User_Predictions'] as $k2 => &$responses)
{
$responses['Answers'] = array_values($responses['Answers']);
}
unset($responses); // break the reference with the last element
}
unset($studyData); // break the reference with the last element
现在数组$studiesData是：
echo "<pre>";
print_r($studiesData);
echo "</pre>";
Array
(
[0] => Array
(
[Study_ID] => 563
[Questions] => Array
(
[0] => S563 Q1
[1] => S563 Q2
)
[User_Predictions] => Array
(
[0] => Array
(
[User_ID] => 24
[Answers] => Array
(
[0] => 66
[1] => Option 1
)
)
)
)
[1] => Array
(
[Study_ID] => 510
[Questions] => Array
(
[0] => S510 Q1
[1] => S510 Q2
[2] => S510 Q3
[3] => S510 Q4
[4] => S510 Q5
)
[User_Predictions] => Array
(
[0] => Array
(
[User_ID] => 76
[Answers] => Array
(
[0] => 1
[1] => 1
[2] => Negative
[3] => 10
[4] => 2,2
)
)
[1] => Array
(
[User_ID] => 58
[Answers] => Array
(
[0] => 25
[1] => 20
[2] => Positive
[3] => 5000
[4] => 1,2
)
)
)
)
)
将所有问题存储在专用数组中，为方便起见，我们将在以后使用它们：
$totalQuestions = array();
foreach ($studiesData as $k => $studyData)
{
foreach ($studyData['User_Predictions'] as $responses)
{
foreach ($studyData['Questions'] as $questionDescription)
{
$totalQuestions[$questionDescription] = $questionDescription;
}
}
}
echo "<pre>";
print_r($totalQuestions);
echo "</pre>";
Array
(
[S563 Q1] => S563 Q1
[S563 Q2] => S563 Q2
[S510 Q1] => S510 Q1
[S510 Q2] => S510 Q2
[S510 Q3] => S510 Q3
[S510 Q4] => S510 Q4
[S510 Q5] => S510 Q5
)
以所需的格式生成结果数组：
$result2 = array();
$result2[0][] = 'User_ID';
$result2[0][] = 'Study_ID';
$x = 1;
foreach($studiesData as $k => $studyData)
{
foreach ($studyData['User_Predictions'] as $k2 => $userData)
{
$x++;
$result2[$x][0] = $userData['User_ID'];
$result2[$x][1] = $studyData['Study_ID'];
foreach($totalQuestions as $questionDescription)
{
if(!in_array($questionDescription, $result2[0]))
{
$result2[0][] = $questionDescription;
}
if(($idQuestion = array_search($questionDescription, $studyData['Questions'])) !== false)
{
$result2[$x][] = $userData['Answers'][$idQuestion];
}
else
{
$result2[$x][] = '';
}
}
}
}
echo "<pre>";
print_r($result2);
echo "</pre>";
最终输出：
Array
(
[0] => Array
(
[0] => User_ID
[1] => Study_ID
[2] => S563 Q1
[3] => S563 Q2
[4] => S510 Q1
[5] => S510 Q2
[6] => S510 Q3
[7] => S510 Q4
[8] => S510 Q5
)
[2] => Array
(
[0] => 24
[1] => 563
[2] => 66
[3] => Option 1
[4] =>
[5] =>
[6] =>
[7] =>
[8] =>
)
[3] => Array
(
[0] => 76
[1] => 510
[2] =>
[3] =>
[4] => 1
[5] => 1
[6] => Negative
[7] => 10
[8] => 2,2
)
[4] => Array
(
[0] => 58
[1] => 510
[2] =>
[3] =>
[4] => 25
[5] => 20
[6] => Positive
[7] => 5000
[8] => 1,2
)
)
一些说明：
array_values返回数组中的所有值，并对数组进行数字索引。
unset在foreach之后：如文档所述：
$ value的引用和最后一个数组元素即使在 foreach循环。建议通过unset（）销毁它。
需要对!==进行false检查，因为此函数可能返回布尔FALSE，但也可能返回非布尔值，其值为FALSE。
Date ：问题的最终输出中有Date条目，但数组输入中没有。如果您更新了问题，我会在答案中包含它。
答案 2 :(得分：0)
我已尝试将其削减，以使步骤尽可能少。
这将其分为两个阶段，收集数据，然后格式化输出。我已在代码中添加了注释，因为它有助于在代码上下文中解释事物...
$entries = [];
$fields = ['User_ID' => null, 'Study_ID' => null];
// Input array of values is $data
foreach ( $data as $study ) {
$studyOptions = [];
foreach ( $study["Questions"] as $question ) {
// Add questions to list of possibilities
$fields[$question] = null;
// Keep a list of the options for this question
$studyOptions[] = $question;
}
foreach ( $study["User_Predictions"] as $prediction ) {
$userData = ["User_ID" => $prediction["User_ID"],
"Study_ID" => $study["Study_ID"]];
// For each prediction, use the corresponding $studyOptions entry to set a field
// with the prediction for this question
foreach ( $prediction["Answers"] as $key => $value ) {
$userData[$studyOptions[$key]] = $value;
}
// Add this to the list of data so far
$entries[] = $userData;
}
}
要显示到目前为止的一些数据...
print_r($entries[0]);
// First row of data example
// Array
// (
// [User_ID] => 24
// [Study_ID] => 563
// [S563 Q1] => 66
// [S563 Q2] => Option 1
// )
print_r($fields);
// List of fields to be used in final output
// Array
// (
// [User_ID] =>
// [Study_ID] =>
// [S563 Q1] =>
// [S563 Q2] =>
// [S510 Q1] =>
// [S510 Q2] =>
// [S510 Q3] =>
// [S510 Q4] =>
// [S510 Q5] =>
// )
最终格式化输出
$output = [array_keys($fields)];
foreach ( $entries as $entry ) {
// Use array_replace - to create an entry with all $fields values
// from the current entry, use array_values to remove keys
$output[] = array_values(array_replace($fields, $entry));
}
给..
Array
(
[0] => Array
(
[0] => User_ID
[1] => Study_ID
[2] => S563 Q1
[3] => S563 Q2
[4] => S510 Q1
[5] => S510 Q2
[6] => S510 Q3
[7] => S510 Q4
[8] => S510 Q5
)
[1] => Array
(
[0] => 24
[1] => 563
[2] => 66
[3] => Option 1
[4] =>
[5] =>
[6] =>
[7] =>
[8] =>
)
....