Home
Limitations of Matplotlib with respect to providing support for customizing plots generated in METviewer, with underlying R-scripting of plots
-
x-tick and y-tick positioning cannot be specified by x- and y-coordinates
-
can only specify the horizontal (left-right) positioning of the plot title via loc='left', loc='right', with default behavior of centered text
-
vertical (up-down) position of the plot title is automatically chosen to avoid labels and ticks on the topmost x-axis (i.e. user cannot specify the distance of the plot title from the topmost x-axis)
-
cannot label the y-axis of the Taylor diagram. This plot is generated with the mpl_toolkits.axisartist.floating_axes.
-
for the Taylor diagram of positive correlation values, points that are close to zero or just below zero correlation (i.e. small negative values) are cut off (unlike the R-script implementation, where these points are visible).
-
cannot modify the location of the y-axis label tick values with respect to the y-axis
-
cannot modify the location of the x-axis label tick values with respect to the x-axis
-
cannot modify the x-axis label title position in the "full" Taylor diagram (i.e. when positive and negative correlation values are plotted). The x-axis label "Standard Deviation" is centered on the x-axis of the positive correlation plot.
Limitations of Plotly with respect to providing support for customizing plots generated in METviewer, with underlying R-scripting of plots
- Axis Margin Line parameters: mgp_title, mgp_labels, mgp_line
- x-axis limits: x1_lim_min, x1_lim_max
- axis values formatting - Perpendicular offset, Frequency, Horizontal align
- add 'show significant' - individual CI intervals can't have custom formatting
- series line types: overplotted, stair steps, histogram-like are not supported
- can't set the box width
- Can't add a title for x2 axis because simple eclv plot would not print all x2 values (too many!). Instead, some of the values are set to '' which moves the x2 axis title above the plot title