Skip to content

CSV Export

WebMACS lets you download all datapoints from a completed experiment as a CSV file — ready to open in Excel, Google Sheets, or import into analysis tools like Python/pandas or MATLAB.

How It Works

The CSV is generated on-the-fly using streaming. This means:

  • No temporary file is created on the server
  • Memory usage stays constant regardless of dataset size
  • Download starts immediately, even for millions of rows
  • Data is ordered by timestamp (oldest first)

Using the UI

  1. Navigate to Experiments in the sidebar
  2. Find a stopped experiment
  3. Click the CSV button
  4. The browser downloads a file named like experiment_Fluidised_Bed_Run_07_a1b2c3d4.csv

Running experiments

You can only export data from stopped experiments. Stop the experiment first if it's still running.

Using the API

curl -o results.csv \
  http://localhost:8000/api/v1/experiments/$EXP_ID/export/csv \
  -H "Authorization: Bearer $TOKEN"

Response Headers

Content-Type: text/csv; charset=utf-8
Content-Disposition: attachment; filename="experiment_My_Run_a1b2c3d4.csv"
Transfer-Encoding: chunked

CSV Format

timestamp,event_name,event_public_id,value,unit,datapoint_public_id
2025-01-15T14:32:10.000000,Inlet Temperature,evt_temp01,23.45,°C,dp_001
2025-01-15T14:32:10.000000,Chamber Pressure,evt_pres01,1.23,bar,dp_002
2025-01-15T14:32:10.500000,Inlet Temperature,evt_temp01,23.51,°C,dp_003

Columns

Column Description
timestamp ISO 8601 timestamp (UTC)
event_name Human-readable sensor/actuator name
event_public_id Event unique identifier
value Measured value (float)
unit Measurement unit (°C, bar, L/min, …)
datapoint_public_id Unique datapoint identifier

Working with the Data

Simply open the downloaded .csv file. Columns are comma-separated with UTF-8 encoding. Most spreadsheet applications detect the format automatically.

import pandas as pd

df = pd.read_csv("experiment_Run_07_a1b2c3d4.csv", parse_dates=["timestamp"])

# Filter by sensor
temp = df[df["event_name"] == "Inlet Temperature"]

# Plot
temp.plot(x="timestamp", y="value", title="Inlet Temperature")

data = readtable('experiment_Run_07_a1b2c3d4.csv');
temp = data(strcmp(data.event_name, 'Inlet Temperature'), :);
plot(temp.timestamp, temp.value);

Tips

Recommendations

  • Export frequently — CSV files are the easiest way to back up experiment data
  • Filter in your analysis tool — the CSV contains all events mixed together; filter by event_name or event_public_id to isolate individual sensors
  • Timestamps are UTC — convert to your local timezone in your analysis tool if needed
  • Large datasets — the streaming approach handles millions of rows, but very large files may take a moment to download

Next Steps