[Qgis-user] (no subject)

[Richard Duivenvoorde]

On 02-01-14 14:14, Willem Korevaar wrote:
> As a geographer teacher I want my students to use Qgis in their spatial
> investigations, using their own data. What I want them to do with GIS is
> to get a rank correlationcoefficient (Rs) using their attribuut data.So
> they have for instance two layers with the same objects (neigbourhoods)
> but with different data ( columns like income and  % shops) . How can
> you use Qgis to get that Rs between income and %shops?
> I can do that in EXcel by copying both columns , rank them and get a
> fomula to get it, but I like to do this directly in Qgis. Is there a way
> (not too complicated I hope)?

Hi Willem,

mmm, rank correlation coefficient... [0] ?

it is not clear to me if you really need two layers. Does your (spatial)
data have both columns income and %shops?
If so, there is no need to load the layer twice, but you can use the
Field Calculator (see [1]) to create/calculate the value via the
creation of your formula using the values of the two columns.

if the data is NOT in the same dataset (eg you have a shp file with
zip-codes and and excel file with columns), you could create a join
first between those two datasets and THEN use the field calculator.

but to be honest, statistics is not my stronges point :-(
so please correct me if I'm wrong.

Regards,

Richard Duienvoorde

[0] http://en.wikipedia.org/wiki/Spearman%27s_rank_correlation_coefficient

[1]
http://qgis.org/en/docs/user_manual/working_with_vector/field_calculator.html