Time Series

The Time Series chart visualises numeric values that change over time. Each numeric series is plotted as a separate line against a shared time axis and is commonly used for prices, metrics, P&L, volumes, and sensor data.

The chart automatically detects time and numeric columns and supports both wide and long data formats. When safe and unambiguous, long-format data is automatically pivoted into aligned series.

Supported Data Formats

1. Wide Format (Already Pivoted)

Each row represents a single point in time and each numeric column represents a separate time series.

1. The first date or time column found is used for the x-axis.
2. Each numeric column becomes one line on the chart.
3. No pivoting or reshaping is required.

dt	seriesA	seriesB
2023-01-01	10.2	12.5
2023-01-02	10.8	12.9

2. Long Format (Time, Series, Value)

Each row represents a single observation for one series at one point in time. This format is common in SQL queries and financial datasets.

dt	sym	value
2023-01-01	A	10.2
2023-01-01	B	12.5
2023-01-02	A	10.8
2023-01-02	B	12.9

When QStudio can unambiguously identify a series dimension, it automatically pivots the data so each series becomes its own line on the chart.

Automatic Time-Series Pivoting

Automatic pivoting reduces boilerplate SQL and allows common time-series queries to work without manual reshaping. It is intentionally conservative to avoid misleading charts.

Conditions Required for Auto-Pivot

All of the following must be true:

1. Exactly one time column (date, timestamp, or time).
2. Exactly one numeric value column.
3. One non-numeric column that:
   • Low cardinality (e.g. symbol, metric)
   • More than one distinct value
   • No duplicate (time, series) combinations
4. The time column is not already unique; if every row has a unique timestamp, no pivot is needed.

Why These Rules Exist

• Ensure there is a single, clear series dimension.
• Prevent accidental pivots when multiple categorical columns exist.
• Avoid silent aggregation or data loss.
• If the structure is ambiguous, the data is rendered exactly as returned.

If auto-pivoting fails for any reason, QStudio safely falls back to the original result set and renders it unchanged.