Mobile edits in 2020

Is there a list of different language versions of Wikipedia by number of edits from mobile?

Paraphrased from memory.

That was a question that piqued my interest in one of the Telegram channels. Since I am a statistics and data fan, I immediately started searching, but I couldn’t find the answer in the otherwise splendid stats.wikimedia.org. There are excellent statistics on page views from desktop versus mobile, but not edits.

Method

I was having a vague memory that I recently saw a similar query in Quarry. After some searching I found that Amir Aharoni already had doing something similar. I made some slight tweaks and had a version that could give you the edits for a full year. By just changing the language code I could then run the query over and over and collect the results in a table. My steam ran out after 136 languages and when the numbers were so small that the proportion varied wildly.

Table of mobile edits in 2020

Below is the table sorted by proportion of edits, if you want a sortable table, check it out on meta. Note that I am missing languages with a “-” in it (like zh-yue). Unfortunately my database skills couldn’t decode the error message. If someone figures out a fix, I’ll be happy to amend the table.

Language codeMobile revisionsTotal revisionsMobile percentage
en66459086236962510.66%
es1628512928378017.54%
ja1179981537537021.95%
it982182771989412.72%
zh855809586646214.59%
fr805308120582276.68%
ko750514282296426.59%
fa571736271640821.05%
ru52080668411227.61%
pt439006290538215.11%
de406647105290483.86%
ar38342198356123.90%
id334876130271125.71%
he28263431901558.86%
vi27809566821914.16%
bn22587685051726.56%
tr18400431563025.83%
pl15529833043884.70%
nl14042724270105.79%
th13132250674225.91%
uk11568734684303.34%
hi11428060192218.99%
el9625362622415.37%
hu9197712774617.20%
az8294268106512.18%
sv8035815616015.15%
ms7536527294027.61%
cs6146112562234.89%
fi556918102516.87%
ro497005500289.04%
sr409337521575.44%
simple408494338519.42%
mr3981313204130.15%
bg365468573474.26%
no344409768123.53%
da307163536638.69%
hr3041930179110.08%
ta2854018623615.32%
ca2819032826710.86%
ur247592678339.24%
ku237098519127.83%
ml2203916712913.19%
uz180947276324.87%
sk1774816725110.61%
hy149338885631.68%
te146962542125.78%
my1419913107210.83%
ckb129718226715.77%
sq1078410259510.51%
kn97845650417.32%
ne97361177238.27%
et91622595643.53%
lv87021832104.75%
as85454421919.32%
eu84817130731.19%
sl70061624514.31%
ga68123655318.64%
tl6709896447.48%
lt65703066612.14%
arz635431890740.20%
km61171229949.74%
kk5574924346.03%
yi53101606333.06%
ka50972940081.73%
ps50431340837.61%
sh4550963784.72%
si43842173020.17%
eo43312673721.62%
gu4258455989.34%
gl38503102631.24%
mk37254684990.80%
bs35881582172.27%
sco3556656615.42%
so3430908337.76%
tg32201097212.93%
ha31881690418.86%
af29592484101.19%
be28832339331.23%
is2859383377.46%
yo24191520615.91%
ast24144574220.53%
or2369279178.49%
pa2322466894.97%
mn2249389545.77%
zu18891453513.00%
sw1786392054.56%
su1553974715.93%
ky15091239812.17%
an1358457312.97%
gom1342703119.09%
tk1312326640.17%
la13061011091.29%
jv1188361923.28%
mzn1144275441.54%
oc1131851251.33%
cy112013240160.08%
als947367532.58%
am942279433.72%
nn9401009960.93%
lij880205584.28%
sah763127565.98%
tcy728107306.78%
ceb6897801120.09%
tt6174451260.14%
lo568425213.36%
min503391861.28%
lb484724650.67%
arc460115239.93%
war434172662.51%
ace40843739.33%
mai39049377.90%
br337313061.08%
wuu308233671.32%
bar306215771.42%
hif29652165.67%
ba284709740.40%
fy274460480.60%
bcl272127522.13%
bjn20135655.64%
ltg19093620.30%
nds107691760.15%
qu10346292.23%
gan9455916.82%
cv88269540.33%
cdo5813064.44%
tet548146.63%
mwl529145.69%
se4140411.01%
iu354387.99%
myv3098450.30%
ady269702.68%
dz212169.72%
ve163035.28%
srn132475.26%
pih112763.99%
ee71644.27%
List of Wikipedia language versions by the number of edits from a mobile phone.

Future work

I would love to see this kind of data in stats.wikimedia.org. If not, finding a way to automize this would also be good, both for completing this dataset with all available languages, but also to do a time series back as far as we have data. However, I am unlikely to be able to do this by myself so if anyone wants to collaborate on it, I would be very happy.