PIN code geocoding

28,326 views
Skip to first unread message

S Anand

unread,
Mar 6, 2011, 1:34:19 PM3/6/11
to Datameet
As discussed in yesterday's datameet, I've set up a crowd-sourced effort to map PIN codes: http://pincode.datameet.org/

It's seeded with 11,000 PIN codes from GeoNames: http://download.geonames.org/export/zip/

Here are ways you can contribute:
1. If you know of any other liberally licensed data sources with the latitude and longitude of PIN codes, let me know (or fork the code and upload it)
2. Even if you had an authoritative list of post codes and names (without lat/lng -- just the list), that'd be of great help too.
3. Any thoughts on how we can merge this with the efforts of hundreds of others doing the same thing are welcome. Hate to overlap effort.

Regards
Anand

Arun Ganesh

unread,
Mar 6, 2011, 2:12:17 PM3/6/11
to data...@googlegroups.com, S Anand
Anand, awesome effort. The largest open source of geocoded pincodes for india are right now sitting in the osm database. It can be extracted using this xapi request:

You will get an osm xml file with all the data, which can also be viewsed visually with josm.
Anand, you can import this data as the base database since it has the majority of the primary pincodes of cities and towns. we can later upload the changes back into osm.
--
j.mp/ArunGanesh

Arun Ganesh

unread,
Mar 6, 2011, 2:20:17 PM3/6/11
to data...@googlegroups.com, S Anand
oops, i gues the xapi service has been down for a few days, i'll see if i can extract the data from my local osm dump and send it.

For reference:
daily osm extracts for india are available from http://download.geofabrik.de/osm/asia/
Data can be selectively extracted from the dump using the tool osmosis
See the osm map features page on what kind of tags are used in the map data
Try out experimental live querying using query-to-map: example of postal codes in chennai
--
j.mp/ArunGanesh

Deepak Shenoy

unread,
Mar 6, 2011, 2:35:09 PM3/6/11
to datameet

On Mar 6, 11:34 pm, S Anand <root.n...@gmail.com> wrote:
> As discussed in yesterday's datameet, I've set up a crowd-sourced effort to
> map PIN codes:http://pincode.datameet.org/
>
> It's seeded with 11,000 PIN codes from GeoNames:http://download.geonames.org/export/zip/

Awesome. How did you do this? It's pretty darn cool = the pincode sub-
site I mean.

http://pincode.net.in/ has 152,000 pin codes, and we could use
wikiscraper or a custom program to get them.

We could run through the pin-code search at http://www.indiapost.gov.in/Pin/Pinsearch.aspx
and get a list of pin codes (gotta do a scraper).

Then use Google Geocoder which has a simple API, just pass
sensor=false&address=[city], India [pincode]. You'll most likely get a
few results, wtih lat/lng, put that into the database, let the crowd
come in and choose the PO directly or choose of of them options.

> 3. Any thoughts on how we can merge this with the efforts of hundreds of
> others doing the same thing are welcome. Hate to overlap effort.

Who else is doing this?

One way to handle overlaps is a) announce it and b) create an api for
people to use (say GetGeoAddress given PinCode). If you know of any of
the overlaps, why don't we list 'em and let's contact everyone and get
the pieces flowing.

Btw, I couldn't be on the skype meet. I'm trying to do stuff to
visualize the budget, am based in Gurgaon, write on financial stuff
and economics in India at www.capitalmind.in, was a programmer in a
past life and am getting back to it, and I constantly work at
visualizing financial data on the blog and privately for investing
ideas. (One of the pieces I'd done was
http://blog.investraction.com/2010/02/bharti-zain-deal-analysis.html -
but that's more or less static.)

S Anand

unread,
Mar 6, 2011, 2:50:54 PM3/6/11
to datameet, Deepak Shenoy
On Sun, Mar 6, 2011 at 7:35 PM, Deepak Shenoy <deepak...@gmail.com> wrote:
Awesome. How did you do this? It's pretty darn cool = the pincode sub-site I mean.

Python, AppEngine, Google Maps API. The source is at https://github.com/sanand0/pincode
 
http://pincode.net.in/ has 152,000 pin codes, and we could use
wikiscraper or a custom program to get them.

Actually, this is a list of 152,000 POST OFFICES, not PIN Codes. There are multiple post offices per PIN code. But you're right, we should be able to get this.

Luckily, this PIN Code Directory is purchasable from http://www.ptcmysore.gov.in/Publication.aspx for Rs 2,000. The other problem is the licensing. I've called them a couple of times, but yet to get into the conversation. If anyone is at Mysore, or knows these folks, would you like a shot at getting this data with a liberal license?
 
Then use Google Geocoder which has a simple API, just pass
sensor=false&address=[city], India [pincode]. You'll most likely get a
few results, wtih lat/lng, put that into the database, let the crowd
come in and choose the PO directly or choose of of them options.

Good point. But re-using the Google Geocoder data will run into licensing problems. I'd rather people did this manually, or we used OpenStreetMap like Arun suggested.
 
One way to handle overlaps is a) announce it and b) create an api for
people to use (say GetGeoAddress given PinCode). If you know of any of
the overlaps, why don't we list 'em and let's contact everyone and get
the pieces flowing.

Don't know who else is doing it. But fair point on the API. Want to give it a shot? Just fork the code and have a play :-)
 
Btw, I couldn't be on the skype meet. I'm trying to do stuff to
visualize the budget, am based in Gurgaon, write on financial stuff
and economics in India at www.capitalmind.in, was a programmer in a
past life and am getting back to it, and I constantly work at
visualizing financial data on the blog and privately for investing
ideas. (One of the pieces I'd done was
http://blog.investraction.com/2010/02/bharti-zain-deal-analysis.html -
but that's more or less static.)

Neat stuff! Tableau is pretty good. In fact, was thinking of having a few tutorials in the next datameet. Would you mind giving a 15 minute tutorial on using Tableau Public?

Thanks
Anand

Arun Ganesh

unread,
Mar 6, 2011, 2:54:11 PM3/6/11
to data...@googlegroups.com, S Anand
I see that we can collect data on two levels pincode geocoding:
1) Mapping of the post office location itself with the associated pincode, but this would be close to impossible in the short term and is more of a long term goal which requires a lot of manpower. Lets not get into this right now.

2) Associating pincodes to place names. This is more approximate and is much more doable. The pincode data in osm is mostly in this form where the point data signifying the location of a city/town/suburb/village has an additional pincode attribute. It might actually be possible to try and automatically geocode a good number of Indian pincodes if we can have an open database of:

A) Geocoded place names. Once again osm is the largest source of such data with cc-by-sa license. Having gone through a lot of osm data, i can say that locations of all major cities and towns are quite accurate. In a few remote areas without road data the positional accuracy drops, sometimes upto ~5km, but otherwise the positions are spot on
Location of villages however are not so good and may have errors of upto ~20km in remote regions. But then again the village data is not very comprehensive and only notable ones are there.

While there are plenty of data sources for geocoded place names, the majority of them are not compatible with the cc-by-sa license because it contains data derived from non free datasets and services like google maps (this may practically be ok, but it is in fact illegal and the derived data belongs to google as per the TOS). Wikipedia and geonames both have such polluted data.
See http://wiki.openstreetmap.org/wiki/Potential_Datasources for a comprehensive listing of usable data sources.

B) Pincode database. This is going to be very tricky because i'm quite sure none of the existing listings on government sites would be compatible with the cc-by-sa license. If we somehow manage to get permission to use the data under this license, it would be possible to match the pincode data with the place names in the osm data and merge them.

***
If the aim is to build a pure creative commons dataset, then you can be rest assured that nobody else has done it for India and this effort is unique and not a duplication of work. Further, to build such a comprehensive database of geocoded pincodes is not an easy task, You may want to have a look at UK's Free the postcode project for an idea of the scale of data collection.
--
j.mp/ArunGanesh

Deepak Shenoy

unread,
Mar 6, 2011, 3:55:35 PM3/6/11
to data...@googlegroups.com
On Mon, Mar 7, 2011 at 1:20 AM, S Anand <root...@gmail.com> wrote:
> On Sun, Mar 6, 2011 at 7:35 PM, Deepak Shenoy <deepak...@gmail.com>
> wrote:

>> Awesome. How did you do this? It's pretty darn cool = the pincode sub-site
>> I mean.
>
> Python, AppEngine, Google Maps API. The source is
> at https://github.com/sanand0/pincode

Absolute coolness. I am learning python, and used version control in
the zamaana where SVN was a relief from Visual Sourcesafe, so please
excuse my ultra-newness to the concept.

>> http://pincode.net.in/ has 152,000 pin codes, and we could use
>> wikiscraper or a custom program to get them.
>
> Actually, this is a list of 152,000 POST OFFICES, not PIN Codes. There are
> multiple post offices per PIN code. But you're right, we should be able to
> get this.

True, even if we got one post office per pin code, and geo tagged it
property it's way more than anything else available! Pincodes are
effectively areas, and I wish we would get a lat/lng polygon but
that's expecting WAAAAY too much from the postal dept right now.

One interesting way to get lat-lang coordinates is from cell phone
towers or have a phone app to directly post that info back into
pincode.datameet.org, for GPS enabled phones. ("What's the Pincode of
your current location" [send]) But it may be overkill.


>> Then use Google Geocoder which has a simple API, just pass
>> sensor=false&address=[city], India [pincode]. You'll most likely get a
>> few results, wtih lat/lng, put that into the database, let the crowd
>> come in and choose the PO directly or choose of of them options.
>
> Good point. But re-using the Google Geocoder data will run into licensing
> problems. I'd rather people did this manually, or we used OpenStreetMap like
> Arun suggested.

Really? Even if we used just the latitude/longitude portions? Quite
interesting.

>> One way to handle overlaps is a) announce it and b) create an api for
>> people to use (say GetGeoAddress given PinCode). If you know of any of
>> the overlaps, why don't we list 'em and let's contact everyone and get
>> the pieces flowing.
>
> Don't know who else is doing it. But fair point on the API. Want to give it
> a shot? Just fork the code and have a play :-)

Will do. Will check out the code too.

> Neat stuff! Tableau is pretty good. In fact, was thinking of having a few
> tutorials in the next datameet. Would you mind giving a 15 minute tutorial
> on using Tableau Public?

Sure - happy to. Whenver we decide on the meet, I need to get a small
dataset together on the concept. It's quite simple, or I haven't used
it enough :)

On a slightly different note: I want to make this for India.
http://www.nytimes.com/interactive/2008/05/03/business/20080403_SPENDING_GRAPHIC.html
I have all the data for India, I know how it's organized etc. But I
have no idea how to start creating a flash (or even Java) based
Voronoi treemap, especially not something that's mapped back into a
circle. Anyone know? Want to collaborate?

Am trying to get into born-again programming, but here's JS source for
on-the-fly voronoi calculations.
http://www.raymondhill.net/voronoi/rhill-voronoi-demo2.php. If you
really like this kinda stuff, you shoudl also know about CGAL.org -
that stuff is a little overhead zone for me at the moment.

Cheers,

Deepak Shenoy
Company: http://www.marketvision.in
Blog: http://www.capitalmind.in
Twitter: @deepakshenoy

S Anand

unread,
Mar 6, 2011, 4:55:19 PM3/6/11
to datameet, Deepak Shenoy
On Sun, Mar 6, 2011 at 8:55 PM, Deepak Shenoy <deepak...@gmail.com> wrote:
On a slightly different note: I want to make this for India.
http://www.nytimes.com/interactive/2008/05/03/business/20080403_SPENDING_GRAPHIC.html
I have all the data for India, I know how it's organized etc. But I
have no idea how to start creating a flash (or even Java) based
Voronoi treemap, especially not something that's mapped back into a
circle. Anyone know? Want to collaborate?

Sure. I was playing around with rectangular treemaps a couple of years ago (http://www.s-anand.net/blog/automating-powerpoint-with-python/), and have been wanting to work on voronoi treemaps for a while. http://www.raymondhill.net/voronoi/rhill-voronoi-core.js looks like a solid implementation. Let me play around with it a bit, and we could start working together on this.

Regards
Anand

Ananth Mani

unread,
Mar 7, 2011, 12:25:20 AM3/7/11
to data...@googlegroups.com, S Anand

@ Anand

Love the speed of implementation - from suggestion to real app. 

I have few more suggestions
  • The pincode displayed in webpage should be editable for entering new pincode. Right now I can do it only through the URL! 
  • Counters - Total Pincode available, Total pincode mapped, Total pincode pending, 10 Randomly picked pincode that needs inputs.
  • I can think more but will restrict to these requests for now :-) 
~ Ananth Mani

Gautam John

unread,
Mar 7, 2011, 1:15:58 AM3/7/11
to datameet
Hey Guys:

This is awesome! A few pointers:

1. Is it possible to add a city field? Much of the trouble seems to be
the lack of one.
2. As part of our work at Akshara, we have collected lat/longs for
almost all Govt. schools in Bangalore along with an address and PIN
Code.

We'd like to give this to the new data meet project and had a question
- we have around 161 unique PIN Codes that represent around 2800
unique institutions with unique locations - do we choose any one
institution that represents a unique PIN Code or is there some geo-
magic you can do with every single lat/long for a unique pin code to
figure out the center point or some such?

Thank you.

Best,

Gautam
________
http://www.aksharafoundation.org/

Arjun Ram

unread,
Mar 7, 2011, 2:00:26 AM3/7/11
to datameet
Hello Folks,

This is a problem that we spend good amount of time on at Taazza. A
couple of quick thoughts on our learnings that you might find useful

- Scraping of pincode site data
We tried this and were largely successfully but for a few issues
with HTTP post. We realized that it was cheaper to buy the data out
for 2K. Happy to opensource the script out. The data should be lying
around somewhere. Let me know.

- Be very cautious with the data from India Post. It isnt upto date
and has loads of typos and has no standards so can be useless.

- Here are some numbers that can be of help.
There needs to be about 30K+ plus pincode as in 560102 (Postal code
offers only 20K odd)
The number of localities within the pincode is lot more - Postal
code only covers 150K Actually numbers are in the millions

- Geonames
The problem is geonames is the data is bad. IMHO, its better to
start off with limited data that is correct rather than wrong data.
May be this would be of help.

- Geocoding of data thru Google
If you intend to make the data available as a DUMP you cannot do
so. License restrictions prevent it.

- Alternate source of data
Yahoo Geoplanet data is good place to start and get folks to geocode
it on top.

In our case we needed this data as a location dictionary for our
semantic engine. You can check for pincodes at http://beta.taazza.com
- use the search box. - Data isnt perfect but a good start.

The million number comes from the fact that the Indian government
actually has the data available and a good number of them are geocoded
as well. Unfortunately they have it in their own system and the values
arent available. The good piece is that the data is available as a
hierarchy which means that if you approach it top down you can iterate
much faster. Arun as far I have heard, has done more work in this and
can probably comment on this.

Also you might want to look at http://linkedgeodata.org/ - There is
lot to be achieved by making this data linked. For instance being able
to query by wikipedia/dbpedia id.

Hope this helps. Let us know if we can be of help.

Best,
----------------------
Arjun Ram
-Founder, Taazza

S Anand

unread,
Mar 7, 2011, 3:01:28 AM3/7/11
to Ananth Mani, datameet
Hi Ananth,
  • The pincode displayed in webpage should be editable for entering new pincode. Right now I can do it only through the URL! 
Done
 
  • Counters - Total Pincode available, Total pincode mapped, Total pincode pending, 10 Randomly picked pincode that needs inputs.

We don't have the full list of PIN codes at the moment, which is part of the problem. But yes, once we have that, this is a definite to-do.

Regards
Anand

Ananth Mani

unread,
Mar 7, 2011, 3:03:59 AM3/7/11
to S Anand, datameet
 
  • Counters - Total Pincode available, Total pincode mapped, Total pincode pending, 10 Randomly picked pincode that needs inputs.

We don't have the full list of PIN codes at the moment, which is part of the problem. But yes, once we have that, this is a definite to-do.

Idea of counter is to reflect the current status (like progress) rather than completed status. 

~ Ananth

Arun Ganesh

unread,
Mar 7, 2011, 5:58:27 AM3/7/11
to data...@googlegroups.com, Ananth Mani, S Anand
Attached a visualization sample of geocoded placenames with pincodes currently in osm. AFAIK this would be the best free datasource currently available.

-Arun
--
j.mp/ArunGanesh
output.png

Dhananjay Nene

unread,
Mar 7, 2011, 6:13:27 AM3/7/11
to data...@googlegroups.com, Arun Ganesh, Ananth Mani, S Anand
How does one generate it for other regions ?

--
-----------------------------------------------------------------------------------
http://blog.dhananjaynene.com twitter: @dnene

Sharun Santhosh

unread,
Mar 7, 2011, 7:20:15 AM3/7/11
to data...@googlegroups.com
Great stuff!

I missed the skype session and am curious what are you guys using this
pincode data for?

Arun Ganesh

unread,
Mar 7, 2011, 7:54:40 AM3/7/11
to Dhananjay Nene, data...@googlegroups.com, Ananth Mani, S Anand
>How does one generate it for other regions ?
 
Check my previous mail on how to download openstreetmap extracts for india and selectively process the data. The visualization i created was generated using a tool called maperitive and a custom stylesheet which displays the postal_code attribute instead of the place name.

A very quick way to do this yourself:
1) Download and unzip the India.osm extract

2) Install the latest version of maperitive and run the following command to load the downloaded osm data into maperitive:
load-source "\foldername\india.osm"

3) Run edit-rules command to edit the current stylesheet and add the following line to the properties section:

properties
    ....
    text:postal_code

4) Save the file and check the maperitive output :) More maperitive help available at http://maperitive.net/docs/manual/

-Arun


On Mon, Mar 7, 2011 at 4:28 PM, Arun Ganesh <arung...@gmail.com> wrote:
> Attached a visualization sample of geocoded placenames with pincodes
> currently in osm. AFAIK this would be the best free datasource currently
> available.
>
> -Arun
> On Mon, Mar 7, 2011 at 1:33 PM, Ananth Mani
> <ananthar...@reportbee.com> wrote:
>>>
>>>
>>>>
>>>> Counters - Total Pincode available, Total pincode mapped, Total pincode
>>>> pending, 10 Randomly picked pincode that needs inputs.
>>>
>>> We don't have the full list of PIN codes at the moment, which is part of
>>> the problem. But yes, once we have that, this is a definite to-do.
>>
>> Idea of counter is to reflect the current status (like progress) rather
>> than completed status.
>> ~ Ananth
>
>
> --
> j.mp/ArunGanesh
>



--
-----------------------------------------------------------------------------------
http://blog.dhananjaynene.com twitter: @dnene



--
j.mp/ArunGanesh

Arun Ganesh

unread,
Mar 7, 2011, 2:54:52 PM3/7/11
to Dhananjay Nene, data...@googlegroups.com, Ananth Mani, S Anand
Attached is the osm xml data of all Indian placenames extracted from the openstreetmap database. I recommend exploring the data and attribute tags using josm. For visualizations like what I created you can play with maperitive. Be warned that the data may have inconsistent tags due to various users contributing data. It can be corrected and updated on the osm server using josm (but be very careful and make sure you know what you are doing).

The data of interest, if you look into the .osm file is the postal_code tag. I have extracted all nodes with a place attribute, and not all of them have the pincode data.

References:
http://wiki.openstreetmap.org/wiki/.osm - osm xml format documentation
http://wiki.openstreetmap.org/wiki/Osmosis - command line utility to process .osm data
--
j.mp/ArunGanesh
india-places.zip

Arun Ganesh

unread,
Mar 7, 2011, 3:06:25 PM3/7/11
to Dhananjay Nene, data...@googlegroups.com, Ananth Mani, S Anand
Forgot to add, most of that data came from a dutch company called AND which donated a lot of their old map datasets to the openstreetmap project. That should explain all the 'AND_nosr" stuff in the xml.
--
j.mp/ArunGanesh

Arun Ganesh

unread,
Mar 7, 2011, 3:18:04 PM3/7/11
to data...@googlegroups.com, Deepak Shenoy

> Good point. But re-using the Google Geocoder data will run into licensing
> problems. I'd rather people did this manually, or we used OpenStreetMap like
> Arun suggested.

Really? Even if we used just the latitude/longitude portions? Quite
interesting.

Using googlemaps to geocode would mean the resulting database is a derivative work of those maps. This breaks 2(b) of the Google Maps Terms of service. One of the biggest databases of crowdsourced map content - wikimapia is locked up because of this. Same is the case with coordinates in wikipedia articles, most of which have been geocoded with a google map based tool.  To avoid license lockdown of the data, its best to stay clear of google maps.

--
j.mp/ArunGanesh

S Anand

unread,
Mar 8, 2011, 7:18:22 AM3/8/11
to Datameet
I've pulled the data out in a format compatible with the AppEngine bulkuploader (osm.csv).

Comparing that with the GeoNames data, we have 2,790 PIN codes in OSM that are new, and I'll import them today. (Arun, this data is CC-SA compatible, right?)

We also 2002 common PIN codes. When I compare the two, some them have significant differences (over 1,000km in some cases). I'm attaching the differences (diff-in-data.csv).

Arun (or anyone else), any guidance on why we have these differences? Which one do you think is more accurate?

Regards
Anand
data.zip

Arun Ganesh

unread,
Mar 8, 2011, 7:50:21 AM3/8/11
to data...@googlegroups.com, S Anand
On Tue, Mar 8, 2011 at 5:48 PM, S Anand <root...@gmail.com> wrote:
I've pulled the data out in a format compatible with the AppEngine bulkuploader (osm.csv).

Comparing that with the GeoNames data, we have 2,790 PIN codes in OSM that are new, and I'll import them today. (Arun, this data is CC-SA compatible, right?)

We also 2002 common PIN codes. When I compare the two, some them have significant differences (over 1,000km in some cases). I'm attaching the differences (diff-in-data.csv).

Arun (or anyone else), any guidance on why we have these differences? Which one do you think is more accurate?

 
A quick glance reveals that many pincodes have the same lat/lon for different pincodes in the geonames data. I think its safe to say that the GN data is much more erroneous than the AND/OSM data. Just sort the cells by the lat values and you'll see what i mean.

I have no idea how AND managed to geocode the pincodes but it seems like they did a pretty good job. I checked 5 random pincodes from the AND data on google and it matched quite closely, And yes, all the openstreetmap data is cc-by-sa 2.0.

-Arun
 
Regards
Anand


On Mon, Mar 7, 2011 at 7:54 PM, Arun Ganesh <arung...@gmail.com> wrote:
Attached is the osm xml data of all Indian placenames extracted from the openstreetmap database. I recommend exploring the data and attribute tags using josm. For visualizations like what I created you can play with maperitive. Be warned that the data may have inconsistent tags due to various users contributing data. It can be corrected and updated on the osm server using josm (but be very careful and make sure you know what you are doing).

The data of interest, if you look into the .osm file is the postal_code tag. I have extracted all nodes with a place attribute, and not all of them have the pincode data.

References:
http://wiki.openstreetmap.org/wiki/.osm - osm xml format documentation
http://wiki.openstreetmap.org/wiki/Osmosis - command line utility to process .osm data
--
j.mp/ArunGanesh




--
j.mp/ArunGanesh

Gautam John

unread,
Mar 8, 2011, 7:54:42 AM3/8/11
to data...@googlegroups.com, Arun Ganesh, S Anand
> google and it matched quite closely, And yes, all the openstreetmap data is
> cc-by-sa 2.0.

Just a point here - the Datameet data set is CC-BY licensed and this
AND is CC-BY-SA licensed. They don't play nicely - you can include AND
in the Datameet set only if the Datameet set is CC-BY-SA licensed too.

S Anand

unread,
Mar 8, 2011, 12:50:32 PM3/8/11
to Gautam John, datameet, Arun Ganesh
Only reason I picked CC-BY is because http://www.geonames.org/ is CC-BY and that's where my initial dataset is from. I can convert that to a CC-BY-SA, right?

Am tempted not to use CC-BY-SA. Attribution is fine, but forcing others to share alike... commercial use would be restricted.

Arun, any chance OSM can be used without ShareAlike?

Regards
Anand

PS: I was just about to upload the 2,000-odd unique PIN codes from datameet when I saw this. Stopped just in time :-)

Arun Ganesh

unread,
Mar 8, 2011, 2:50:29 PM3/8/11
to S Anand, Gautam John, datameet


On Tue, Mar 8, 2011 at 11:20 PM, S Anand <root...@gmail.com> wrote:
Only reason I picked CC-BY is because http://www.geonames.org/ is CC-BY and that's where my initial dataset is from. I can convert that to a CC-BY-SA, right?

Am tempted not to use CC-BY-SA. Attribution is fine, but forcing others to share alike... commercial use would be restricted.

Arun, any chance OSM can be used without ShareAlike?

I'm no expert on legal matters, if you ask me, i'll point you to OSM License Page and Gautam :)
--
j.mp/ArunGanesh

Gautam John

unread,
Mar 9, 2011, 12:42:57 AM3/9/11
to Arun Ganesh, S Anand, datameet
Sadly, I can think of a way around this easily. Some commentary here:
http://www.osmfoundation.org/wiki/License/Why_CC_BY-SA_is_Unsuitable

The one way around this is to keep the OSM data set independent of
everything else, display it independently and return changes to OSM as
mandated by the SA license. The other option, is to ask AND if they
can independently license it to data meet under a vanilla BY license.

chaitanya reddy

unread,
Mar 25, 2015, 1:08:46 AM3/25/15
to data...@googlegroups.com
Hi

I might be replying to the old thread. But useful information. We can find the all the pin codes of India here: https://data.gov.in/sites/default/files/all_india_pin_code.csv

Thank you
Chaitanya

Shivaprakash Yaragal

unread,
Jan 30, 2016, 7:03:56 AM1/30/16
to datameet
we need coordinates chaitanya; they dont have they have only pincodes.

Dilip Damle

unread,
Feb 1, 2016, 8:01:38 AM2/1/16
to datameet
HI, 
I think I shared this pincode data earlier once when I attended the pincode meet in Bangalore. 
It is a CSV file with about 29000 unique pincodes with about 2lakh location names within those pincodes
I had collected it from various online sources.
See if it is of any help. NO LAT LON here.
Some of them may be deprecated.
All_india_pincodes_combined_data_1.7z

Sharad Bhat

unread,
Aug 30, 2016, 9:26:13 AM8/30/16
to datameet
I recently came across a list of 1.5 lakh location which have been geo-coded. Pin code data is also in the file, and there are ~19,000 unique pincodes - https://dl.dropboxusercontent.com/u/62163696/India%20pin%20codes%20lat%20lon%20map.xlsb

Devdatta Tengshe

unread,
Aug 30, 2016, 12:35:32 PM8/30/16
to data...@googlegroups.com
While this data looks extensive, I have serious reservations about it's accuracy.

1) You can see that there are many points which fall outside the boundary of the country

2) I checked for a few known pincodes, and the points are placed hundreds of km away.

3) I checked the first 2 digits which indicate the sorting district, and there are many points which appear in totally different regions of the country,

​Use this data at your own risk.

Regards,
Devdatta

--
Datameet is a community of Data Science enthusiasts in India. Know more about us by visiting http://datameet.org
---
You received this message because you are subscribed to the Google Groups "datameet" group.
To unsubscribe from this group and stop receiving emails from it, send an email to datameet+unsubscribe@googlegroups.com.
For more options, visit https://groups.google.com/d/optout.

Sandeep Singh

unread,
May 10, 2018, 9:49:35 AM5/10/18
to datameet
Hi Sharad,

The link is not accessible.

Regards,
Sandeep

Sharad Bhat

unread,
May 10, 2018, 11:12:51 AM5/10/18
to data...@googlegroups.com
Hi Sandeep,

I've removed it, as it was not very accurate, but you should be able to find an alternative with a little googling

--
Datameet is a community of Data Science enthusiasts in India. Know more about us by visiting http://datameet.org
---
You received this message because you are subscribed to a topic in the Google Groups "datameet" group.
To unsubscribe from this topic, visit https://groups.google.com/d/topic/datameet/21uKP0AzrjY/unsubscribe.
To unsubscribe from this group and all its topics, send an email to datameet+unsubscribe@googlegroups.com.

For more options, visit https://groups.google.com/d/optout.



--
Regards,
Sharad Bhat

yamraj....@gmail.com

unread,
May 22, 2018, 9:14:28 AM5/22/18
to datameet
On Monday, 7 March 2011 00:04:19 UTC+5:30, S Anand wrote:
> As discussed in yesterday's datameet, I've set up a crowd-sourced effort to map PIN codes: http://pincode.datameet.org/
>
>
> It's seeded with 11,000 PIN codes from GeoNames: http://download.geonames.org/export/zip/
>
>
>
> Here are ways you can contribute:
> 1. If you know of any other liberally licensed data sources with the latitude and longitude of PIN codes, let me know (or fork the code and upload it)
>
> 2. Even if you had an authoritative list of post codes and names (without lat/lng -- just the list), that'd be of great help too.
> 3. Any thoughts on how we can merge this with the efforts of hundreds of others doing the same thing are welcome. Hate to overlap effort.
>
>
>
> Regards
> Anand

https://groups.google.com/forum/#!forum/india-pincode--population-data

Atul Rawle

unread,
Aug 10, 2020, 8:26:05 AM8/10/20
to datameet
Can you share Again?

apurva...@gmail.com

unread,
May 24, 2021, 5:48:59 PM5/24/21
to datameet
cant Your share this data again 
Reply all
Reply to author
Forward
0 new messages