# Summary Turns marked activities into numbers: mean concentration per task, and exposure metrics against occupational limits. This is usually the tab you came for — the one whose table ends up in the report. ```{figure} ../_static/gui/tab-summary.png :alt: The Summary tab showing a per-activity table with duration, mean and standard deviation for each dataset :width: 100% One row per dataset × activity, with the statistics you asked for. Several instruments are combined into a single table. ``` **Available for:** any project. It works with one dataset or many — tick several and they are combined into one table with `Dataset` and `Instrument` columns, so you can compare instruments task by task. Mark your activities on the [Time series](time-series.md) tab first; with none marked you only get an **All data** row. ## Datasets to include Tick the datasets to summarise. Nothing is computed until you press **Compute**, because exposure statistics over several long datasets can take a moment. ## Type **Activity summary** — descriptive statistics per activity: duration plus whichever of **Mean**, **Std**, **Min**, **Max** and **Median** you tick. **Exposure summary** — the occupational-hygiene view: the time-weighted average over the TWA window, the highest short-term average, peaks, and how each compares against the limits you set. It reveals the extra fields: | Field | Meaning | | --- | --- | | **STEL (short-term limit)** | Short-term exposure limit. The highest short-window average is compared against it. | | **over** | Averaging window for the STEL check, as a pandas offset — `15min` by default. | | **OEL (8h limit)** | Occupational exposure limit. The time-weighted average is compared against it. | | **TWA window** | Averaging window for the time-weighted average — `8h` by default. | Both limits are in the **same unit as the metric you chose**, so if you are summarising a mass concentration in µg/m³, enter the limit in µg/m³. ## Choose metrics… Picks which quantities to summarise, grouped by instrument. Two things make this work across a mixed project: - Each metric is computed **only for the datasets that provide it**, so adding a gas monitor to a project full of particle counters does not fill the table with blanks. - Comparable quantities recorded at different unit scales — ng/m³ and µg/m³, say — are merged into one column. ## Compute, and the staleness warning **Compute** builds the table. The result is cached and saved with the project, so reopening it later shows what you computed without recalculating. Because it is cached, it can go out of date. If you edit activities, change the data, or alter any of the limits, an amber banner appears: > ⚠ These values may be out of date — tasks, data, or settings changed since they > were computed. Click Compute to refresh. Take it seriously — the numbers shown were computed from the *previous* inputs. ## Export to Excel… Writes the combined table to `.xlsx` or `.csv`, exactly as displayed. ## Under the hood ```python import aerosoltools as at data = at.load_ops_file("measurement.csv") # Activity summary data.summarize_activities(stats=["mean", "std"]) # Exposure summary data.summarize_exposure( metric="PM4.2", long_limit=1.0, # OEL (8h limit) twa_window="8h", # TWA window short_limit=1.0, # STEL short_window="15min", # over ) ``` See [4 — Statistics and exposure](../examples/04-statistics-and-exposure.ipynb).