I'm a little stumped how to visualize this in a meaningful way.
Here it is again with the latest data. This is the best visual I think out
of the few I tried:
http://www-958.ibm.com/software/analytics/manyeyes/visualizations/aoo-downloads-download-density
On Thu, Jan 24, 2013 at 10:28 AM, Rob Weir <[email protected]> wrote:
On Wed, Jan 23, 2013 at 7:16 PM, Kadal Amutham <[email protected]> wrote:
Dear Rob,
For the internet population I used the data from the wiki since the
link
was already in the google spreadsheet. For the population details, I
typed
population followed by the country name in the address bar of google
and
in
most of the searches the result was from www.google.co.in/publicdata.
I
do
not expect much difference between google data and wiki. You can verify
few
data and the difference may be in the last 3 digits.
I reviewed the data. Most of it looked fine. As you say, maybe in
the last few digits. I only found a few things to fix. For example
(and I didn't know this before) there are two Congo's now: Democratic
Republic of the Congo and the Republic of the Congo. It looks like the
data there was reversed. Little things like that.
The top 30 countries, for downloading AOO, normalized by number of
internet users in the country, are:
1. Monaco
2. Italy
3. France
4. San Marino
5. Luxembourg
6. Belgium
7. Estonia
8. Germany
9. Switzerland
10. Finland
11. Spain
12. Hong Kong
13. Austria
14. Netherlands
15. Canada
16. Belize
17. Czech Republic
18. Falkland Islands (Malvinas)
19. Faroe Islands
20. Malta
21. Singapore
22. Iceland
23. Taiwan
24. United Kingdom
25. United States
26. Uruguay
27. Hungary
28. New Zealand
29. Ireland
30. Liechtenstein
I'm attaching the full spreadsheet. Hopefully Samer can put the
updated figures on the ManEyes website.
The countries at the bottom of the list, are some where we have
opportunities for growth. For example, South Korea should improve
once the Korean translation is released next week.
-Rob
With Warm Regards
V.Kadal Amutham
919444360480
914422396480
On 23 January 2013 23:58, Samer Mansour <[email protected]> wrote:
Some countries didn't get counted (not matched in the tool/didn't
exist).
Some cities got their own row, but the Map doesn't have a spot for
them. ie
Hong Kong and Singapore:
Also I think some rows are gone, like "Europe (not specified)", it was
there yesterday.
Here are two more visualizations:
http://www-958.ibm.com/software/analytics/manyeyes/visualizations/apache-openoffice-downloads-vs-pop
http://www-958.ibm.com/software/analytics/manyeyes/visualizations/apache-openoffice-downloads-vs-pop-2
And the original:
http://www-958.ibm.com/software/analytics/manyeyes/visualizations/apache-openoffice-downloads-by-pop
I didn't have a design in mind in how we want to show this so I just
plotted and mashed around with it like playdough.
On Wed, Jan 23, 2013 at 12:37 PM, Rob Weir <[email protected]>
wrote:
On Wed, Jan 23, 2013 at 10:59 AM, Samer Mansour <[email protected]
wrote:
Opps here it is:
http://www-958.ibm.com/software/analytics/manyeyes/visualizations/apache-openoffice-downloads-by-pop
This is very nice. We might want to take out the "population" and
"internet users" and the two rank columns and just use the
"downloads
per 1k population" and "downloads per 1k internet users" columns.
I found a few cut & paste errors in the spreadsheet, so it would be
good if we could have some reviewers double check the numbers.
Compare columns C and E of this spreadsheet:
https://docs.google.com/spreadsheet/ccc?key=0Av4Lhq3W5zKodGZtNU1oRGFjWi1kYXkzVEtjOWY1ZlE#gid=0
With the data in these two Wikipedia articles.
http://en.wikipedia.org/wiki/List_of_countries_by_number_of_Internet_users
http://en.wikipedia.org/wiki/List_of_countries_by_population
Once reviewed, and we have the updated maps, then I have an idea. We
could do a blog post on these numbers, but frame it as a general
story
about end-user use of desktop open source software, etc. If we make
the story more general interest we'll get broader circulation and
uptake in the press. Of course, it supports a positive story of our
project as well. But by telling the broader story, and using
OpenOffice as an example, we'll go further.
Does this make sense?
-Rob
On Wed, Jan 23, 2013 at 10:59 AM, Samer Mansour <
[email protected]
wrote:
Here is an updated map. I moved to bubble chart because coloring
the
country made places like Russia and Canada look odd.
I will see if I can make other visualizations as well, maybe Map
is
not
the best visualization.
On Wed, Jan 23, 2013 at 3:10 AM, Kadal Amutham <[email protected]
wrote:
Dear Rob Weir,
The internet user data has been filled
With Warm Regards
V.Kadal Amutham
919444360480
914422396480
On 23 January 2013 07:23, Kadal Amutham <[email protected]>
wrote:
I will start filling the internet user data
With Warm Regards
V.Kadal Amutham
919444360480
914422396480
On 23 January 2013 01:38, Rob Weir <[email protected]>
wrote:
On Tue, Jan 22, 2013 at 12:11 PM, Kadal Amutham <
[email protected]
wrote:
Check the google document at
https://docs.google.com/spreadsheet/ccc?key=0Av4Lhq3W5zKodGZtNU1oRGFjWi1kYXkzVEtjOWY1ZlE#gid=0
I have entered the population data alone
Very good. I made some edits and also received a spreadsheet
with
some
more data from Gianvittorio.
The results are interesting. For example, what countries
have
the
highest percentage of downloads by population? The top 5
are:
1) Gambia
2) Trinidad and Tobago
3) Zimbabwe
4) Vatican City
5) Saint Pierre and Miquelon
It will be interesting to see how the data is when we look at
usage
per internet users in a country.
But however we slice the data we'll have this problem: For
very
small
countries, just a few downloads can shift the ratio by a
large
amount.
For example, the Vatican City had around 100 downloads. So
a
difference of 10 downloads is 10%. France had 4.5 million
downloads.
A difference of 10 downloads is nothing.
I've seen this handled in other contexts by applying
statistical
techniques to estimate error bounds on the ratio, and then
rank by
the
lower confidence limit. You can see the technique described
(and
formulas given) here:
http://evanmiller.org/how-not-to-sort-by-average-rating.html
-Rob
With Warm Regards
V.Kadal Amutham
919444360480
914422396480
On 22 January 2013 22:26, Kadal Amutham <[email protected]>
wrote:
Good show. I have almost entered the population data into
the
spread
sheet. Then I have to populate the internet user data.
Once
it
is
over, I
will let you know.
With Warm Regards
V.Kadal Amutham
919444360480
914422396480
On 22 January 2013 21:37, Samer Mansour <
[email protected]
wrote:
We don't have internet users completely populated so from
the
data
we
currently have I whipped something up in 5 mins at work.
This is just the top 40 Countries. Its manual data
entry,
once
we
have
internet users populated and I will resolve things like
"Hong
Kong" ->
"China" and Singapore -> South Korea.
I can maybe finish some data entry tonight.
http://www-958.ibm.com/software/analytics/manyeyes/visualizations/aoo-test-visualization
This is a PNG output:
http://www-958.ibm.com/software/analytics/manyeyes/vis/FullScreen/fullscreenvisualization.html?id=files%2Fthumbnails%2Ffed90050-64ac-11e2-926b-000255111976.wm.png&visId=ff062cd864ac11e2926b000255111976
On Mon, Jan 21, 2013 at 7:44 AM, Rob Weir <
[email protected]
wrote:
On Sun, Jan 20, 2013 at 10:34 PM, Kadal Amutham <
[email protected]>
wrote:
Dear Mr. Rob, I tried to edit the document, but it is
read
only.
Can
you
make it readable so that everybody can fill the data?
OK. I gave you and Samer write permissions.
-Rob
With Warm Regards
V.Kadal Amutham
919444360480
914422396480
On 21 January 2013 09:01, Kadal Amutham <
[email protected]
wrote:
Who ever finds time, can fill the remaining Data, so
that
the
document
become complete
With Warm Regards
V.Kadal Amutham
919444360480
914422396480
On 21 January 2013 07:46, Rob Weir <
[email protected]>
wrote:
On Sat, Jan 19, 2013 at 12:34 PM, Roberto
Galoppini <
[email protected]>
wrote:
On Fri, Jan 18, 2013 at 10:34 PM, Rob Weir <
[email protected]>
wrote:
I was thinking of putting together a world map
showing
the
use of
OpenOffice, the kind with each country shaded or
color
coded
to
show
the density of use.
I can easily get a data set showing the total
number
of
downloads of
OpenOffice per country. But the raw numbers
don't
really
tell
the
story. It would show, probably, that the USA
has
the
most
downloads.
But that is probably also because of its large
population.
So maybe we then show downloads per capita, or
downloads
per
100,000
population. But that then becomes a proxy for
economic
development,
since there are highly populated countries with
fewer
computers
per
capital, and low population countries with more
computers,
etc.
I
don't think that is what we want to show.
So, I'm wondering, has anyone seen data for
something
like
PCs
per
capita, or home computers, or internet users, or
some
other
proxy
for
what our potential usership would be per
country?
Maybe we can use internet users stats by country,
something
like:
http://en.wikipedia.org/wiki/List_of_countries_by_number_of_Internet_usersto
normalize our stats.
Thanks, that looks useful. I started entering the
data
into
a
spreadsheet:
https://docs.google.com/spreadsheet/ccc?key=0Av4Lhq3W5zKodGZtNU1oRGFjWi1kYXkzVEtjOWY1ZlE
As you see, Italy is at the top, if you look at
downloads
per
1000
internet users, with 91. So nearly one in ten in
Italy
have
downloaded AOO!
-Rob
Regards,
-Rob
--
====
This e- mail message is intended only for the
named
recipient(s)
above.
It
may contain confidential and privileged
information.
If
you
are
not
the
intended recipient you are hereby notified that
any
dissemination,
distribution or copying of this e-mail and any
attachment(s)
is
strictly
prohibited. If you have received this e-mail in
error,
please
immediately
notify the sender by replying to this e-mail and
delete
the
message
and
any
attachment(s) from your system. Thank you.
