From 280d36230052de4f94e384648c1283031fbc9840 Mon Sep 17 00:00:00 2001 From: Johannes Ranke Date: Tue, 17 Jul 2018 17:29:14 +0200 Subject: Fix inverse predictions for replicate measurements For details, see NEWS.md --- docs/news/index.html | 153 +++++++++++++++++++++++++++++++++++++++++++++++++++ 1 file changed, 153 insertions(+) create mode 100644 docs/news/index.html (limited to 'docs/news/index.html') diff --git a/docs/news/index.html b/docs/news/index.html new file mode 100644 index 0000000..62ab315 --- /dev/null +++ b/docs/news/index.html @@ -0,0 +1,153 @@ + + + + + + + + +Changelog • chemCal + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + + +
+
+ + + +
+ +
+
+
+

Changelog

+ +
+ +
+

+chemCal 0.2.1 (2018-07-17) Unreleased +

+
    +

  • ‘inverse.predict’: Do not work on the means of the calibration standards any more, as this ignores the variability of y values about the means

    • +

  • Use testthat for tests to simplify further development. Adapt the tests using data with replicate standard measurements to work on the means in order to show the relation to ‘inverse.predict’ from earlier versions. Include comparisons with investr::calibrate(method = ‘Wald’) for unweighted regressions. Include tests with more precision to check for changes in numerical output across versions.

    • +

  • ‘lod’ and ‘loq’: In the lists that are returned, return the list component ‘y’ without names, because we always only have a single element in ‘y’ (previously the name ‘1’ was returned).

    • +

  • Convert vignette to html and explain the changes to ‘inverse.predict’

    • +

  • Add two example dataset, one from an online course at the University of Toronto, one from Rocke and Lorenzato (1995)

    • +
+
+
+

+chemCal 0.1-33 (2014-04-24) Unreleased +

+
    +
  • Bugfix in lod() and loq(): In the case of small absolute x values (e.g. on the order of 1e-4 and smaller), the lod or loq calculated using the default method could produce inaccurate results as the default tolerance that was used in the internal call to optimize is inappropriate in such cases. Now a reasonable default is used which can be overriden by the user. Thanks to Jérôme Ambroise for reporting the bug.
    • +
+

For a detailed list of changes to the chemCal source please consult the commit history on https://cgit.jrwb.de/chemCal

+
+
+ + + +
+ + +
+ + + + + + -- cgit v1.2.1