Skip to content

Latest commit

 

History

History
130 lines (77 loc) · 5.34 KB

vindta.md

File metadata and controls

130 lines (77 loc) · 5.34 KB
Raw

VINDTA DIC processing

koolstof only deals with dissolved inorganic carbon. For total alkalinity, use Calkulate.

Like Calkulate, koolstof works by adding extra methods and attributes to a standard pandas DataFrame, which contains a .dbs file generated by a VINDTA. The corresponding VINDTA-generated logfile.bak is also required.

All the code examples assume the import convention:

import koolstof as ks

Import VINDTA files

The logfile.bak: read_logfile()

Import a logfile.bak into a standard pandas DataFrame with one row per DIC sample.

logfile = ks.read_logfile("path/to/logfile.bak", methods="3C standard")

!!! example "read_logfile: optional keyword arguments"

* `methods`: list of VINDTA method filenames used to run samples, excluding the `.mth` extensions

The .dbs file: read_dbs()

Import a .dbs file into an enhanced DataFrame, rename its columns into a friendlier format, and link it to the already imported logfile DataFrame.

dbs = ks.read_dbs("path/to/file.dbs", keep_all_cols=False, logfile=None)

!!! example "read_dbs: optional keyword arguments"

* `keep_all_cols`: retain all columns from the `.dbs` (`True`) or just the most important ones (`False`)?
* `logfile`: the DataFrame generated by importing the `logfile.bak` with `read_logfile()`.

Add sample metadata

Once you've imported the files above, you need to add the following metadata as extra columns in the dbs enhanced DataFrame under the following column labels:

  • salinity: practical salinity (assumed 35 if not provided)
  • temperature_analysis_dic: temperature of DIC analysis in °C (assumed 25 °C if not provided)
  • dic_certified: certified DIC values for reference materials in μmol/kg-sw. Non-reference samples should be set to np.nan.

You can also add the following logical columns to refine which samples and reference materials are used for processing and calibration:

  • blank_good: use this sample in assessing the coulometer blank value?
  • k_dic_good: use this reference material for calibration?

Calibrate DIC measurements

Once you have imported the logfile.bak, .dbs file and added metadata, the dbs enhanced DataFrame has a series of methods available to calibrate the DIC results.

These must be run in a specific order, but each one also checks whether the necessary step before it has been run, and runs it automatically if not. Therefore you only actually need to call the final method dbs.calibrate_dic(), unless for some reason you need to intervene during the processing.

!!! warning "Previous steps are only carried out if necessary"

The later processing steps only re-run the earlier steps if they are required.  Mostly this is determined by checking for the presence or absence of a specific column in the `dbs` DataFrame, as described below.

!!! example "All processing steps in order"

The full sequence of steps, plus their keyword arguments, is:

1. [`dbs.get_logfile_index()`](#dbsget_logfile_index)
2. [`dbs.get_sample_blanks(use_from=6)`](#dbsget_sample_blanks)
3. [`dbs.get_session_blanks(batch_col="dic_cell_id")`](#dbsget_session_blanks)
4. [`dbs.get_blank_corrections()`](#dbsget_blank_corrections)
5. [`dbs.get_density()`](#dbsget_density)
6. [`dbs.get_standard_calibrations()`](#dbsget_standard_calibrations)
7. [`dbs.get_session_calibrations(batch_col="dic_cell_id")`](#dbsget_session_calibrations)
8. [`dbs.calibrate_dic()`](#dbscalibrate_dic)

See below for details of each step.

!!! tip "Keyword arguments are passed along"

If using one of the later functions, you can pass any keyword arguments needed by the earlier functions to it as well, and they will get passed along until they are needed.

For example, if you do `dbs.calibrate_dic(use_from=5)`, then the `use_from=5` gets passed along to the `dbs.get_sample_blanks()` step.

dbs.get_logfile_index()

dbs.get_logfile_index()

Finds the row in the logfile corresponding to each row of the dbs based on matching both the sample name in dbs["bottle"] and the analysis date and time.

Adds a column to dbs:

  • logfile_index: the index for the row in

dbs.get_sample_blanks()

dbs.get_sample_blanks(use_from=6)

Checks for a logfile_index column in dbs, and runs dbs.get_logfile_index() if absent.

Estimates the coulometer blank for each sample as the average of the count increments, starting at the number of minutes set by use_from.

Adds a column to dbs:

  • blank_here: the coulometer blank for each sample in counts per minute.

dbs.get_session_blanks()

dbs.get_session_blanks(batch_col="dic_cell_id")

Checks for a blank_here column in dbs, and runs dbs.get_sample_blanks() if absent.

Fits the coulometer blanks for each sample as a function of time of analysis for each analysis session (as identified by the column dbs[batch_col]).

Adds an attribute to dbs:

  • sessions: a standard pandas DataFrame containing one row of metadata per analysis session.

Visualise the calibrations

dbs.plot_blanks()

dbs.plot_k_dic()

dbs.plot_dic_offset()