Input Timeseries Data

To run crspy-lite, you need timeseries data of your sensors "counting rate". You also need timeseries data of several meteorological variables for the processing, in order to form correction factors, calibrate the sensor and perform quality control. See below for guidance on the formating of this input data:

Required Columns

The table below summarises the required variables that comprise the input data, denoting the expected column name and unit. Please note that all columns must be the same length. Any gaps in the data should be filled with 'nan' values.

NAME	UNITS	DESCRIPTION
DATETIME	Datetime	Datetime of the observation. Format: "yyyy-mm-dd hh:mm:ss"
MOD	Count	Moderated neutron count for time interval.
PRESS	mb	Pressure sensor measurement.
TEMP	Celsius	External Temperature at the site. This would be an external reading.
E_RH	(%)	External Relative Humidity at the site.
I_RH	(%)	Relative humidity inside the sensor box.
BATT	Volts	Voltage of the CRNS battery.
RAIN	mm	Rainfall at the site.

Automatic formatting

Crspy-lite will automatically format your input data in several ways. This includes:

• Detecting the delimiter of your input data and reading it correspondingly.
• Resampling your input data to an hourly temporal resolution.
• Fixing the dataframe to values "on the hour" (e.g. 10:00 instead of 10:13).
• Flipping the dataframe if the dates are the wrong way around.
• Removing whitespace in columns and drop any empty ones (e.g. columns with only NaN values in them).

Structuring the input data

Crspy-lite is dependent on the user to structure their input data correctly and upload it via either a filepath or a url to the config.json file. Check the table below for a formatting checklist:

Filepath	URL
Included all required columns	Included all required columns
Correct column names	Correct column names
File is .txt or .csv with a consistent delimiter	Consistent delimiter (tab or comma recommended)
Ensure the filepath sting is correctly formatted	Ensure the link is publically accessible (no authentication required)
	Ensure the link points directly to the file of the data

For clarification of these formatting requirements click on the links for examples of input timeseries data via a filepath and a URL respectively. You can also see the example of the config file to check how the structure of the filepath string.