Normalisation of data

Question

I am trying to plot the data below in a pie chart. I split the pie chart based on the group first and then based on the Id. But since for some rows, the count is very small, I am not able to see it in the pie chart.

I am trying to normalise the data. I am not sure on how to do that. Any help would be sincerely appreciated.

 Group Id Count G1 12 276938 G1 13 102 G2 12 27 G3 12 4683 G3 13 7 G4 12 301

You could try charting the log of the count. What's your end goal? — Mohammad Athar
– Mohammad Athar, Commented Sep 12, 2016 at 17:05
My end goal is to normalise the above data in a specific range. — user3447653
– user3447653, Commented Sep 12, 2016 at 17:06
what I mean is why do you want a pie chart? any linear normalization won't change the size of the slices; did you try charting the log of the count? — Mohammad Athar
– Mohammad Athar, Commented Sep 12, 2016 at 17:07
Normalisation means scaling amounts measured with different scales. It doesn't apply to these data since they are measured on the same scale. A pie chart is not appropriate way of representing these data. Instead a bar-chart with a logarithmic scale is possible, but may be deceptive, or separate G1-12 from the others — James K
– James K, Commented Sep 12, 2016 at 17:12
@user2539738: No I have not tried it still. Is taking a log of count a normalisation technique. — user3447653
– user3447653, Commented Sep 12, 2016 at 17:12

Zeugma · Accepted Answer · 2016-09-12 17:14:13Z

Don't pie chart what doesn't fit a visual representation in a pie chart

( df.groupby(['Group', 'Id']) .sum().Count.sort_values(ascending=False) .plot.bar(logy=True, subplots=True) )

Collectives™ on Stack Overflow

Normalisation of data

1 Answer 1

Comments

Hot Network Questions

Collectives™ on Stack Overflow

1 Answer 1

Comments

Related