{
  "title": "Autocorrelation diagnostic",
  "human_url": "https://happytails.ai/en/calculators/autocorrelation-diagnostic/",
  "agent_url": "https://happytails.ai/agents/en/calculators/autocorrelation-diagnostic/",
  "language": "en",
  "markdown_url": "https://happytails.ai/en/calculators/autocorrelation-diagnostic.md",
  "content": "Autocorrelation diagnostic\nInspect lag dependence with an autocorrelation chart, pointwise white-noise references and explicit independence limitations.\nHow can we improve this tool?\nReport a problem or share an idea.\nPlease leave out private information. Your message and the tool name go to our team via PostHog.\nHow to use Autocorrelation diagnostic\nPaste complete numeric CSV rows with the required headers. Preserve the original order for lag diagnostics.\nChoose the displayed method settings and review their assumptions.\nRead the chart and table, then download all observations, results and assumptions as JSON or CSV.\nUseful next steps\nEmpirical distribution viewer\nConfidence interval calculator\nBox plot comparison\nAutocorrelation diagnostic\nChoose a whole maximum lag from 1 to the smaller of 100 and the number of rows minus one. Use a value column in observation order with equally spaced measurements. The lag calculation uses the full centered sum of squares as its denominator. Constant series have undefined autocorrelation. Dashed bands are approximate pointwise 95% white-noise references, not simultaneous tests. Dependence can invalidate independent-sample standard errors; an effective sample size is not inferred without a model. No detected autocorrelation does not prove independence or randomness.\nKeep the complete data trail\nSearch and sort change this comparison only. Copy result copies the summary. Download JSON or CSV for all observations, results, settings and method assumptions. Download SVG for the chart. Group labels are preserved. Numeric CSV inputs use a decimal point; editable numeric settings also accept the selected language’s decimal notation.",
  "content_format": "text/plain",
  "visibility": "public",
  "links": [
    {
      "name": "Empirical distribution viewer",
      "url": "https://happytails.ai/en/calculators/empirical-distribution-viewer/"
    },
    {
      "name": "Confidence interval calculator",
      "url": "https://happytails.ai/en/calculators/confidence-interval-calculator/"
    },
    {
      "name": "Box plot comparison",
      "url": "https://happytails.ai/en/calculators/box-plot-comparison/"
    }
  ],
  "tool": {
    "id": "autocorrelation-diagnostic",
    "category": "calculators",
    "built": true,
    "requirements": "Paste comma-separated CSV with a header and 2 to 2000 complete rows. Use dot decimals with at most six decimal places and values from -1000000000 to 1000000000. No missing values are silently dropped. Extra columns are used only to preserve the original input in reports. Calculations use the required columns only. Processing stays on your device. Results use floating-point arithmetic. Charts show the supplied sample, not proof about a population. Choose a whole maximum lag from 1 to the smaller of 100 and the number of rows minus one. Use a value column in observation order with equally spaced measurements. The lag calculation uses the full centered sum of squares as its denominator. Constant series have undefined autocorrelation. Dashed bands are approximate pointwise 95% white-noise references, not simultaneous tests. Dependence can invalidate independent-sample standard errors; an effective sample size is not inferred without a model. No detected autocorrelation does not prove independence or randomness. Search and sort change this comparison only. Copy result copies the summary. Download JSON or CSV for all observations, results, settings and method assumptions. Download SVG for the chart. One disposable worker per calculation, with a 10-second timeout. CSV supports up to 100 columns and 100000 cells through the shared parser.",
    "implementation": {
      "fields": [
        {
          "key": "maxLag",
          "label": "Maximum lag",
          "value": "1",
          "type": "text",
          "inputMode": "numeric",
          "maxLength": 24
        }
      ],
      "presentation": "metrics",
      "inputMaxLength": 1000000,
      "inputErrorCode": "DIAGNOSTIC_INPUT",
      "note": "Paste comma-separated CSV with a header and 2 to 2000 complete rows. Use dot decimals with at most six decimal places and values from -1000000000 to 1000000000. No missing values are silently dropped. Extra columns are used only to preserve the original input in reports. Calculations use the required columns only. Processing stays on your device. Results use floating-point arithmetic. Charts show the supplied sample, not proof about a population. Choose a whole maximum lag from 1 to the smaller of 100 and the number of rows minus one. Use a value column in observation order with equally spaced measurements. The lag calculation uses the full centered sum of squares as its denominator. Constant series have undefined autocorrelation. Dashed bands are approximate pointwise 95% white-noise references, not simultaneous tests. Dependence can invalidate independent-sample standard errors; an effective sample size is not inferred without a model. No detected autocorrelation does not prove independence or randomness.",
      "experience": {
        "automatic": false,
        "single": false,
        "inputMaxLength": 1000000,
        "inputMode": "text",
        "label": "Numeric CSV",
        "placeholder": "value\n1\n2\n3\n4",
        "help": "Paste comma-separated CSV with a header and 2 to 2000 complete rows. Use dot decimals with at most six decimal places and values from -1000000000 to 1000000000. No missing values are silently dropped. Extra columns are used only to preserve the original input in reports. Calculations use the required columns only. Processing stays on your device. Results use floating-point arithmetic. Charts show the supplied sample, not proof about a population. Choose a whole maximum lag from 1 to the smaller of 100 and the number of rows minus one. Use a value column in observation order with equally spaced measurements. The lag calculation uses the full centered sum of squares as its denominator. Constant series have undefined autocorrelation. Dashed bands are approximate pointwise 95% white-noise references, not simultaneous tests. Dependence can invalidate independent-sample standard errors; an effective sample size is not inferred without a model. No detected autocorrelation does not prove independence or randomness. Up to 1,000,000 characters. Review the settings before running.",
        "action": null
      },
      "mode": "explicit",
      "interactive": false
    },
    "execution": "Browser interface",
    "api": null,
    "browser_agent": {
      "supported": true,
      "name": "run_current_tool",
      "discovery": "WebMCP on the human page in a supporting browser",
      "verification": "See audit/API-AUDIT.md; availability is not verification"
    }
  },
  "markdown": "# Autocorrelation diagnostic\n\n- Human page: https://happytails.ai/en/calculators/autocorrelation-diagnostic/\n- Markdown: https://happytails.ai/en/calculators/autocorrelation-diagnostic.md\n- Structured JSON: https://happytails.ai/agents/en/calculators/autocorrelation-diagnostic/index.json\n- Visibility: public\n\n## Tool capabilities\n\n| Property | Value |\n| --- | --- |\n| Tool ID | `autocorrelation-diagnostic` |\n| Category | calculators |\n| Implementation | Registered implementation. Registration alone is not a production-quality guarantee. |\n| Browser execution | Use the visible action or interactive controls. |\n| Browser processing | Browser; see tool notes for network behavior. |\n| HTTP API | Not callable through HTTP. |\n| Browser-agent interface | run_current_tool |\n\n### Implementation notes\n\nPaste comma-separated CSV with a header and 2 to 2000 complete rows. Use dot decimals with at most six decimal places and values from -1000000000 to 1000000000. No missing values are silently dropped. Extra columns are used only to preserve the original input in reports. Calculations use the required columns only. Processing stays on your device. Results use floating-point arithmetic. Charts show the supplied sample, not proof about a population. Choose a whole maximum lag from 1 to the smaller of 100 and the number of rows minus one. Use a value column in observation order with equally spaced measurements. The lag calculation uses the full centered sum of squares as its denominator. Constant series have undefined autocorrelation. Dashed bands are approximate pointwise 95% white-noise references, not simultaneous tests. Dependence can invalidate independent-sample standard errors; an effective sample size is not inferred without a model. No detected autocorrelation does not prove independence or randomness.\n\n### Inputs and settings\n\nText input: Numeric CSV.\n\nExample or placeholder shown in the interface (not a validated fixture):\n\n```text\nvalue\n1\n2\n3\n4\n```\n\nPaste comma-separated CSV with a header and 2 to 2000 complete rows. Use dot decimals with at most six decimal places and values from -1000000000 to 1000000000. No missing values are silently dropped. Extra columns are used only to preserve the original input in reports. Calculations use the required columns only. Processing stays on your device. Results use floating-point arithmetic. Charts show the supplied sample, not proof about a population. Choose a whole maximum lag from 1 to the smaller of 100 and the number of rows minus one. Use a value column in observation order with equally spaced measurements. The lag calculation uses the full centered sum of squares as its denominator. Constant series have undefined autocorrelation. Dashed bands are approximate pointwise 95% white-noise references, not simultaneous tests. Dependence can invalidate independent-sample standard errors; an effective sample size is not inferred without a model. No detected autocorrelation does not prove independence or randomness. Up to 1,000,000 characters. Review the settings before running.\n\n| Setting | Meaning | Control type | Default | Constraints |\n| --- | --- | --- | --- | --- |\n| `maxLag` | Maximum lag | text | \"1\" |  |\n\n### Browser-agent workflow\n\n1. Open the human page in a browser that supports WebMCP.\n2. Discover the interface exposed by that page; use its actual schema.\n3. Select files on the page first when the tool requires files.\n4. Call `run_current_tool` with valid input and settings.\n5. Inspect the returned result or error and the displayed output. Retrieve files from the displayed download links.\n\nCalls do not automatically copy or download results. Browser permissions still apply. Interface availability alone is not proof of successful execution.\n\n### Planning specification\n\nPaste comma-separated CSV with a header and 2 to 2000 complete rows. Use dot decimals with at most six decimal places and values from -1000000000 to 1000000000. No missing values are silently dropped. Extra columns are used only to preserve the original input in reports. Calculations use the required columns only. Processing stays on your device. Results use floating-point arithmetic. Charts show the supplied sample, not proof about a population. Choose a whole maximum lag from 1 to the smaller of 100 and the number of rows minus one. Use a value column in observation order with equally spaced measurements. The lag calculation uses the full centered sum of squares as its denominator. Constant series have undefined autocorrelation. Dashed bands are approximate pointwise 95% white-noise references, not simultaneous tests. Dependence can invalidate independent-sample standard errors; an effective sample size is not inferred without a model. No detected autocorrelation does not prove independence or randomness. Search and sort change this comparison only. Copy result copies the summary. Download JSON or CSV for all observations, results, settings and method assumptions. Download SVG for the chart. One disposable worker per calculation, with a 10-second timeout. CSV supports up to 100 columns and 100000 cells through the shared parser.\n\nThis describes intended requirements. Use the implementation notes, interface schema and observed output to determine current support.\n\n### Complete machine-readable capability record\n\n```json\n{\n  \"id\": \"autocorrelation-diagnostic\",\n  \"category\": \"calculators\",\n  \"built\": true,\n  \"requirements\": \"Paste comma-separated CSV with a header and 2 to 2000 complete rows. Use dot decimals with at most six decimal places and values from -1000000000 to 1000000000. No missing values are silently dropped. Extra columns are used only to preserve the original input in reports. Calculations use the required columns only. Processing stays on your device. Results use floating-point arithmetic. Charts show the supplied sample, not proof about a population. Choose a whole maximum lag from 1 to the smaller of 100 and the number of rows minus one. Use a value column in observation order with equally spaced measurements. The lag calculation uses the full centered sum of squares as its denominator. Constant series have undefined autocorrelation. Dashed bands are approximate pointwise 95% white-noise references, not simultaneous tests. Dependence can invalidate independent-sample standard errors; an effective sample size is not inferred without a model. No detected autocorrelation does not prove independence or randomness. Search and sort change this comparison only. Copy result copies the summary. Download JSON or CSV for all observations, results, settings and method assumptions. Download SVG for the chart. One disposable worker per calculation, with a 10-second timeout. CSV supports up to 100 columns and 100000 cells through the shared parser.\",\n  \"implementation\": {\n    \"fields\": [\n      {\n        \"key\": \"maxLag\",\n        \"label\": \"Maximum lag\",\n        \"value\": \"1\",\n        \"type\": \"text\",\n        \"inputMode\": \"numeric\",\n        \"maxLength\": 24\n      }\n    ],\n    \"presentation\": \"metrics\",\n    \"inputMaxLength\": 1000000,\n    \"inputErrorCode\": \"DIAGNOSTIC_INPUT\",\n    \"note\": \"Paste comma-separated CSV with a header and 2 to 2000 complete rows. Use dot decimals with at most six decimal places and values from -1000000000 to 1000000000. No missing values are silently dropped. Extra columns are used only to preserve the original input in reports. Calculations use the required columns only. Processing stays on your device. Results use floating-point arithmetic. Charts show the supplied sample, not proof about a population. Choose a whole maximum lag from 1 to the smaller of 100 and the number of rows minus one. Use a value column in observation order with equally spaced measurements. The lag calculation uses the full centered sum of squares as its denominator. Constant series have undefined autocorrelation. Dashed bands are approximate pointwise 95% white-noise references, not simultaneous tests. Dependence can invalidate independent-sample standard errors; an effective sample size is not inferred without a model. No detected autocorrelation does not prove independence or randomness.\",\n    \"experience\": {\n      \"automatic\": false,\n      \"single\": false,\n      \"inputMaxLength\": 1000000,\n      \"inputMode\": \"text\",\n      \"label\": \"Numeric CSV\",\n      \"placeholder\": \"value\\n1\\n2\\n3\\n4\",\n      \"help\": \"Paste comma-separated CSV with a header and 2 to 2000 complete rows. Use dot decimals with at most six decimal places and values from -1000000000 to 1000000000. No missing values are silently dropped. Extra columns are used only to preserve the original input in reports. Calculations use the required columns only. Processing stays on your device. Results use floating-point arithmetic. Charts show the supplied sample, not proof about a population. Choose a whole maximum lag from 1 to the smaller of 100 and the number of rows minus one. Use a value column in observation order with equally spaced measurements. The lag calculation uses the full centered sum of squares as its denominator. Constant series have undefined autocorrelation. Dashed bands are approximate pointwise 95% white-noise references, not simultaneous tests. Dependence can invalidate independent-sample standard errors; an effective sample size is not inferred without a model. No detected autocorrelation does not prove independence or randomness. Up to 1,000,000 characters. Review the settings before running.\",\n      \"action\": null\n    },\n    \"mode\": \"explicit\",\n    \"interactive\": false\n  },\n  \"execution\": \"Browser interface\",\n  \"api\": null,\n  \"browser_agent\": {\n    \"supported\": true,\n    \"name\": \"run_current_tool\",\n    \"discovery\": \"WebMCP on the human page in a supporting browser\",\n    \"verification\": \"See audit/API-AUDIT.md; availability is not verification\"\n  }\n}\n```\n\n## Page guide\n\nInspect lag dependence with an autocorrelation chart, pointwise white-noise references and explicit independence limitations.\n\nHow can we improve this tool?\n\nReport a problem or share an idea.\n\nPlease leave out private information. Your message and the tool name go to our team via PostHog.\n\n### How to use Autocorrelation diagnostic\n\n1. Paste complete numeric CSV rows with the required headers. Preserve the original order for lag diagnostics.\n2. Choose the displayed method settings and review their assumptions.\n3. Read the chart and table, then download all observations, results and assumptions as JSON or CSV.\n\n## Useful next steps\n\n- [Empirical distribution viewer](https://happytails.ai/en/calculators/empirical-distribution-viewer/) ([Markdown](https://happytails.ai/en/calculators/empirical-distribution-viewer.md))\n\n- [Confidence interval calculator](https://happytails.ai/en/calculators/confidence-interval-calculator/) ([Markdown](https://happytails.ai/en/calculators/confidence-interval-calculator.md))\n\n- [Box plot comparison](https://happytails.ai/en/calculators/box-plot-comparison/) ([Markdown](https://happytails.ai/en/calculators/box-plot-comparison.md))\n\n## Autocorrelation diagnostic\n\nChoose a whole maximum lag from 1 to the smaller of 100 and the number of rows minus one. Use a value column in observation order with equally spaced measurements. The lag calculation uses the full centered sum of squares as its denominator. Constant series have undefined autocorrelation. Dashed bands are approximate pointwise 95% white-noise references, not simultaneous tests. Dependence can invalidate independent-sample standard errors; an effective sample size is not inferred without a model. No detected autocorrelation does not prove independence or randomness.\n\n### Keep the complete data trail\n\nSearch and sort change this comparison only. Copy result copies the summary. Download JSON or CSV for all observations, results, settings and method assumptions. Download SVG for the chart. Group labels are preserved. Numeric CSV inputs use a decimal point; editable numeric settings also accept the selected language’s decimal notation.\n"
}