microsoft / vivainsights Goto Github PK

For validation report: While using pq_data under validation_report in vivainsights error is,

Error in rename():
! Can't rename columns that don't exist.
✖ Column Instant_Message_hours doesn't exist.

Kindly assist for sample data to use validation_report in vivainsights. Thanks

Ability to easily customize the node color scheme, e.g. by applying a palette,

e.g. as in ggraph

custom_palette <- c("red", "blue", "green", "yellow")... ggraph(graph) + ... scale_color_manual(values = custom_palette) + ...

Ability to change node shape,

e.g. in ggraph to denote IC v. manager:

geom_node_point(aes(shape = TieOrigin_supervisor_indicator))...

It might be cool to have a base function that users can take to generate doughnut visuals, like the following:

This visual would be especially useful to be embedded in a dashboard/report output.

This idea is inspired by the following post:
https://stackoverflow.com/questions/74569982/ggplot-create-dynamic-multi-kpis-donut-in-one-chart

Noticed the last subplot of extreme right's X axis positioned above the rest of the subplots X axis positions.

Ability to calculate node clustering coefficient, in a similar way as centralities

e.g. as in igraph:

cluster_coef <- transitivity(g, type="local")

When we specify the intended bins range to the functions - afterhours_dist(), collaboration_dist() and create_hist(), colour for the last two bins range is same, default bins is working perfect.

Please refer the below examples.

Ex:1 Same intensity of red colour displayed for 25 - 35 hours and 35 + hours range

Ex:2 Same intensity of red colour displayed for 10 - 15 hours and 15 + hours range, so not able to visualise the data

Ex:3 Default bins, no issue .

Currently, the plot output of create_line() assumes that the vertical axis will represent 'Weekly hours', as it is explicitly stated in the code.

However, this is not necessarily the case, as this can be used to measure variables like internal network size.

A more appropriate value would be the metric value that is provided to the parameter of the function, ideally formatted with us_to_space() so that underscores are replaced with spaces.

Ability to easily use a few public domain HR datasets so that users can try out the functions— the random graphs don’t work as well because they don’t seem to be distributed like a professional org would

e.g. as in 3.13 here: https://ona-book.org/viz-graphs.html

Noticed difficulty in visualising the data in Trend plots due to the used diverging color map. Please refer below plots.

Currently, PersonId and MetricDate may not be in the exact spelling or case when exported from a localized Analyst experience.

Therefore, adding the ability in import_query() to rename these columns could enable versatility when such a scenario arises.