Keep the measurement comparable
Use the same measure, definition, location, and same unit throughout: monthly means and daily maxima answer different questions. Check comparable dates, source coverage, and changes in measurement method.
Spiral can help you explore repeating patterns in a time series. Start with a defined cycle, comparable observations, and a clear question; wrapping time does not establish a seasonal effect or its cause.
For example: does one location show a recurring pattern in monthly temperature over several years? Start with that question and reviewed observations; no dataset or result is included here.
Use the same measure, definition, location, and same unit throughout: monthly means and daily maxima answer different questions. Check comparable dates, source coverage, and changes in measurement method.
For a year of monthly observations, choose Monthly: 12 points per turn. Review one observation per consecutive month and its date. Granularity sets the number of points around a turn; it does not aggregate daily rows into monthly means or repair gaps.
Check gaps and duplicate dates: unknown is not zero. Explain incomplete cycles and any preparation. Animation between observations adds no measurements, and seasonal patterns do not establish causality.
The available choices are Hourly (24 points), Daily (365), Monthly (12), Yearly (10), and None (all timeline points in one turn). With recognized dates, the first date provides a starting offset; later angles advance by point order. Irregular intervals, missing months, leap days, or time-zone changes can shift a comparison. Check the actual dates and positions; there is no arbitrary custom cycle control or automatic seasonal alignment guarantee.
Prepare a stable entity name, date or period, and numeric value for each observation, plus its source. For one location, repeat its name. Use consistently formatted dates, such as YYYY-MM-DD, and review chronological order. Choose New chart → Upload a file when ready; sign in if asked.
Prepare and import your source table →Use the existing request guide's Metric over time path. Specify the entity, measure, unit, date range, source, and observation frequency. Ask for gaps and preparation to be explained before making a cycle comparison.
Generation and applying data are voluntary and may use existing AI credits. Review the proposal and original sources. This page does not submit a request, generate or import data, or select Spiral for you.
Write a request & review the native steps →For sharing, follow the native format, Universal safe zone, and export guide. Use Export → Export Chart and open the downloaded MP4 at phone size. Free exports retain the watermark. These instructions do not certify a Spiral video; each composition needs its own Preview and received-file checks.
Best for
Choose a native points-per-turn setting and check that it matches your prepared observation frequency.
Use Trail Length to choose how much history is retained and inspect incomplete cycles before drawing a comparison.
Color can encode metric value, trail age, or absolute time. State that choice rather than treating color as evidence of a cause.
Workflow
Use one measure and unit, consistent dates, and an explicit cycle. Review gaps, duplicate rows, and changes in source coverage before import.
Confirm chronological Fields mapping, choose Chart → Chart Type → Spiral, and set native Granularity, trail, labels, grid, and colors.
Check cycle positions and phone-size readability in Preview, wait for Saved, and open the received MP4 before sharing.
Questions
ChartGlow keeps the workflow simple while leaving the visual and data controls in your hands.
No. Hourly uses 24 points per turn, Daily 365, Monthly 12, Yearly 10, and None uses all timeline points in one turn. A recognized first date supplies a starting offset, but later angles advance by point index. The setting does not aggregate data, validate frequency, or repair irregular intervals. Check gaps, leap days, time zones, and the actual positions; there is no arbitrary custom-cycle control.
Use consistent calendar labels such as YYYY, YYYY-MM, or YYYY-MM-DD. In Fields, select Time labels → Auto (calendar dates when recognized), map Time Column or Order Source, and use Order → Ascending. If labels are not all recognized, sorting can fall back to numeric or alphabetical order. Formatting a label does not validate its underlying date or frequency.
The native renderer supports multiple entities on the timeline. Visible series depend on filtering and Max Items; importing several entities does not guarantee all appear at every frame. Start with one series, then use comparable units and reviewed periods if adding others. Recheck your own Preview and MP4; this page does not qualify a multi-series video.
Unknown is not zero. Blank or non-numeric metric cells can become zero; duplicate rows for an entity and date are added. Resolve these before import and explain any preparation. Review Fill Missing Values: carrying a previous value forward adds an assumption, not a measurement. Cumulative Sum changes the metric. Missing timeline periods can shift later cycle positions even when a gap is not filled.
No. A recurring visual pattern is a question to investigate with source coverage, measurement definitions, and other evidence. Animation between observations adds no measurements. Neither seasonality nor a color scale establishes causality.
Explore other formats
Follow changing ranks among selected entities with animated bars, names, values, and images.
Show trajectories, crossovers, momentum, and long-term change with animated lines.
Compare changing magnitudes with animated circles, values, flags, and images.