Version 0.8.3

• Correctly format "Academic Period" for the summer session.

Version 0.8.2

Changed

• Removed grading basis column since it is no longer supported.
• Change "read more" link to point to the new documentation

Version 0.8.1

• Fixed Windows-specific error with format string for timestamps

Version 0.8.0

Added

• xlsxwriter as a dependency (for Workday formatting for grades).

Changed

• BREAKING Changed the CSV export format from FSC to Workday.
• Updated mentions of "FSC" to be more general, including the file names.
• Clarified error message for ungraded assignments.

Fixed

• Print fewer columns for students with an override score so that it is easier to overview.
• Silenced a pandas future warning and restricted from upgrading to to 3.0.

Version 0.7.0

Added

• Warn when there are ungraded assignments, similar to the warning for unposted assignments.

Fixed

• Show less rows and columns in warnings to make the output easier to overview.
• Fix numexpr issue with macs not being able to use the C engine for pandas query

Version 0.6.2

Fixed

• Switch to check "creation" instead of "start" date in the show-courses subcommand,
  since many courses don't have a "start" date listed.
• Fix handling of NAs in date strings for show-courses subcommand.
• Check that all necessary fields exist and raise an informative error if they do not.

Version 0.6.1

Fixed

• Fix regression introduced in 0.6.0 where the checking for the student id was done incorrectly so that it always evaluated to False.

Version 0.6.0

Added

• Automatically detect if there are multiple sections and display a boxplot for each section automatically.
• Include an option flag to include whether to group by section, grader, or nothing for the charts.
• Add search box to filter students names in the charts.
• Add percentile information to each student in the hover info in the scatter plot.
• Show progress bar while downloading grades
• Check for students being part of multiple sections.

Changed

• Use boxplots with hover info (count, mean, median, quartiles) instead of just a marker for mean and median. Useful to report both overall stats and for each assignment/quiz.
• Extract student specific section ID instead of propagating the same ID to all students. Useful for courses with combined sections on Canvas.
• Lay out the scatterplot as a violin cloud so that it is easier to tell the shape of the distribution and the layout is deterministic when changing the dropdown menu or filtering via the search box.
• Prompt for inclusion of assignments and make the regex easier to use.
• Change the default number of histogram bins and the range to be more suitable for most grading scenarios.

Fixed

• Adjust position of text and widgets to make the charts clearer.
• Support missing student numbers (happens with concluded courses).
• Support section IDs that are not numeric
• Align scales and ticks between boxplots and histograms.

Version 0.5.1

Changed

• Remove placeholders for commands already included in the viz

Version 0.5.0

Added

• Visualize individual student assignment scores.
  • This makes it possible to see how a student is doing over time
    and if intervention is needed.
• Optionally filter how many assignments show up in the visualization.
• Link Final Grade Plot with individual student assignment scores.
• Visualize score distributions for each assignment along with mean and median.
• Visualize a comparison between scores from each grader.
• Center chart titles and add more elaborate instructions.
• Show mean and median grade in plot.

Changed

• Require Python 3.8 instead of 3.6 to ensure that importlib.metadata is in stdlib.
• Make saved messages stand out and simplify final info note.
• Make hover highlighting more intuitive and better explained.
• Filter out unpublished assignment and those missing a max score.
• Put assignment titles on top of plots instead of to the left.
• Maintain assignment order from Canvas in all visualizations.
• Only show a sample of the students with unposted assignments.

Fixed

• Make rounding work with dfs containing None instead of NaN
  (happens when all values are None for an assignment and there is no type casting)