search

kibichii / geoda

Automatically exported from code.google.com/p/geoda
GNU General Public License v3.0
0 stars 0 forks source link

Percentile Map: Add values to categories in Legend #21

Open GoogleCodeExporter opened 9 years ago

GoogleCodeExporter commented 9 years ago 
from  Sep 29, 2011

This comes from Richard on the OpenSpace list.  Here's the original suggestion:

When making a percentile map it would be useful to be able to see what the 
actual values are for the data in the legend, so if I see 10%-50% (528) (the 
percentile and the number of polygons in that percentile range, being able to 
see the data value boundaries, say 10%-50% (528) 33.45-67.01  would be very 
helpful.

I'd also suggest that we add the ability to click on legend categories to 
select all observations in that particular category.  A user could then use the 
Table to see what the category boundaries are.

Original issue reported on code.google.com by mmcc...@gmail.com on 5 Mar 2013 at 6:06

GoogleCodeExporter commented 9 years ago 
There is currently a logical inconsistency between the labels for the 
percentile map and the labels for the other maps. The percentile map does not 
show the bounds of the categories, the others do. This needs to be changed to 
conform.

Original comment by lanse...@gmail.com on 25 Jun 2015 at 2:36

GoogleCodeExporter commented 9 years ago 
propose a solution for this fix, see the attached figure (first). The second 
figure seems too crowd.

the real values are appended at the end of the label.  

Did not use the style "10%-50% (528) 33.45-67.01", since it's too crowded (see 
second figure)

Let me know any comments.

Original comment by lixun...@gmail.com on 2 Jul 2015 at 9:25

Attachments:

GoogleCodeExporter commented 9 years ago 
fixed in ci #3680

Original comment by lixun...@gmail.com on 2 Jul 2015 at 9:56

GoogleCodeExporter commented 9 years ago 
fix will be applied in 1.7.37

change status for verification

Original comment by lixun...@gmail.com on 9 Jul 2015 at 6:44

GoogleCodeExporter commented 9 years ago 
Hi Xun, Please add the minimum and maximum values to the first and last legend 
item so people know what those are (e.g. this is often relevant for quality 
checks of the data to make sure there aren't any numbers that are outside of 
the expected value range). Thx, Julia

Original comment by jkoschin...@gmail.com on 13 Jul 2015 at 5:13

GoogleCodeExporter commented 9 years ago 
fixed in ci 3073

Julia, can you check the value range of percentile? looks weird me -- why <1% 
and 1%-10% has 0 items?

Original comment by lixun...@gmail.com on 14 Jul 2015 at 2:32

GoogleCodeExporter commented 9 years ago 
This is related to the value distribution of the HR60 variable (homicides in 
1960), which has a lot of zeros. Natural breaks would be a better map 
classification since there's no way to uniquely cut the zeros into the three 
bottom categories, so they're all put into the 10-50% bin.

Original comment by jkoschin...@gmail.com on 14 Jul 2015 at 9:03

GoogleCodeExporter commented 9 years ago

Original comment by jkoschin...@gmail.com on 14 Jul 2015 at 1:12

GoogleCodeExporter commented 9 years ago 
One other small cosmetic thing: Could you set the precision of the legend 
values to two after the decimal point, e.g. 2.34? Right now the natural breaks 
map has 6 digits, and equal intervals+quantile 3 or two (see attached). thx!

Original comment by jkoschin...@gmail.com on 16 Jul 2015 at 11:22

Attachments:

GoogleCodeExporter commented 9 years ago 
Luc discovered one other issue with the boxmap legend (see attached): When no 
observations are in the lower outlier category, the minimum value gets set to 
zero instead of the actual minimum value of the variable. E.g. in attached min 
value is zero but should be -0.17.

Original comment by jkoschin...@gmail.com on 16 Jul 2015 at 11:40

GoogleCodeExporter commented 9 years ago 
Luc discovered one other issue with the boxmap legend (see attached): When no 
observations are in the lower outlier category, the minimum value should be set 
to “:” or something like that since there are no values in that range. Same 
on the high end, if the max value is less than the upper fence, the upper end 
value should be “:” not the max.

Original comment by jkoschin...@gmail.com on 16 Jul 2015 at 11:44

Attachments:

GoogleCodeExporter commented 9 years ago 
Mark to accepted for fixing

Original comment by lixun...@gmail.com on 16 Jul 2015 at 8:08

GoogleCodeExporter commented 9 years ago 
Fix has been proposed in ci 3721 3722

Fix will be in >= 1.7.45

Original comment by lixun...@gmail.com on 21 Jul 2015 at 1:25

GoogleCodeExporter commented 9 years ago 
Let's skip the scientific notation unless the numbers are over 100k (e.g see 
attached screenshot of SIDS data). When we used scientific notation before, we 
got lots of emails from users who can't read it.

Could you add a conditional statement for when the numbers are very small (e.g. 
0.001 as in the attached SIDS example) to go up to 4 digits after the decimal 
point since the range is otherwise meaningless (as in the attached example with 
0.00 everywhere). If it's beyond 4, we can switch to scientific notation.

Original comment by jkoschin...@gmail.com on 27 Jul 2015 at 12:30

Attachments: