Is there a way to create a custom colour palette with specific colours? Right now, we can only specify one or more colour families in the custom palette section in "Data Colours" and Omniscope generates a palette. What if I know the exact RGB values of the colours I want in the palette?
All colour pickers in Omniscope should allow you to enter RGB values directly including pasting the #-prefixed hex/web codes.
In "Style > Data colours > Edit category colours", you can customise the auto colouring of categories, but if you want to specify a particular colour for a particular value, make sure the field is a Category field and see "Data > Manage fields > [field] > Options > Value order, colours and shapes".
For continuous data (numerics, dates and graded categories), in "Style > Data colours > Edit continuous colours" you can set the start/mid/end colours precisely, but cannot colour specific values without first converting to Categories.
Thanks for your suggestions. I am aware of the category colors. There are some inefficiencies around manually assigning colours to category values. Examples would be:
1. A random (not really random, but definitely not preconfigured) colour would be assigned to a new category that gets added to the data set. 2. If I have a lot of category values, and I need my charts to have only approved corporate colors, then it is a lot of work to set the colors manually to category values. Also, when I try to automate the reporting process, I will always have a risk of a new category coming in and a "non-corporate" colour or shade getting applied to it.
The solution I propose is this:
An option to create "Custom" palette from the "Data Colors" menu, where I can restrict what are the "valid" colours for my IOK file.
I've reclassified this as an idea, so people can vote.
Please clarify exactly how you would specify corporate colours while still avoiding the same colour being assigned to several values. Do you anticipate a series of permitted corporate hues being supplied, while Omniscope creates variations in brightness?
You can experiment with this on a single colour/hue basis using "Style > Data colours > Edit category colours > Multiple shades of a single base colour". I'm guessing you want exactly the same behaviour, but being able to specify multiple "permitted hues" or "base colours"?
Exactly Steve. The way it works on most cases in our experience, corporate branding defines primary colours (2-3 colors) and secondary colours (5 or 6 colours). Also, organisations differ in the way they are strict about branding. Some may allow variations in brightness, while others will only allow specific RGB values.
So, ideally, we should be able to add multiple "hues" to a palette and then an option saying "Generate shades" or some such. I hope this helps?
Simply allowing multiple base colours will resolve this issue. We'll implement this subject to demand/votes.
However, be warned, Omniscope assigns colours to values globally. So "Xyz" is always green/whatever, everywhere it appears. This means that, if you have only 5 approved colours, and brightness variations are not allowed, you will very frequently see the same colours in a given bar chart, often next to each other.
Unlikely without more votes or justification. A feature which allows you to restrict the colour palette to 5 colours is not very useful for data visualisation, regardless of corporate policy.
Data colours are typically used to correlate the same value across multiple views or to avoid labels and use a colour key.
Let's say a typical pie view has 10 segments, and you have 5 colours. You already can't identify these accurately in the colour key or across multiple views. And should additional values appear in the incoming data, this makes the problem even worse.
If you instead want to use colours for non-data elements such as toolbars or to colour all bars/pies/etc. (using the "default data colour" shown when not using colour-by-value), this should be configurable without being affected by changes in the incoming data.
Note that the colour picker for categories has a "memory" of the last 20 or so colours chosen, so it's easy to reuse colours precisely.
If I am missing a use case or have misunderstood your request, please correct me.
I'll describe the scenario that prompted me to put in this idea.
I have a client for whom I'm building dashboards. A typical IOK report uses around 50 category fields. Some of these category fields have 50-100 unique values.
If I have to manually assign colours, I'll have to repeat the color grabbing exercise a few thousand times. which is not very efficient. Also, with there is a problem with this approach. If I get a new value in a category field, then it can be a colour which I have not manually chosen.
The branding requirement is this:
We have 2 main colors, and 5 secondary colors. We can use different variations of the 2 main colors and exact shade of 5 secondary colors.
Currently you can choose variations of a single colour in the data colours options. I suggest you try this with one of the 2 main colours, and perhaps use shades of the other for numerics/dates.
To make full use of your corporate colour restrictions, you would either need some extended custom colour palette which specifies exactly those rules you've described - this is pretty specific - or you would need a colour palette which allowed you to define a "pool" of available colours. But you'd need to define quite a few shades to avoid getting too many clashes, and I'm not sure given these 5+2 restrictions you'd be able to pull this off.