abresler / realtr

Real estate brokers are the....

realtR

Installation

devtools::install_github("hadley/emo")
devtools::install_github("abresler/realtR")

realtR

There are few service industries worse than residential real estate brokerage industry.

Industry actors are some of the most overpaid, dishonest, people you will encounter.

They hoard information, lie, are rarely held accountable for anything they do. They are a well organized political cartel who ensure regardless of what happens, they get paid a sizable chunk of the overall proceeds from a transaction.

The consumer has a hard time educating themselves and can be at the mercy of these snake-oil salesmen.

The time has come to change for and realtR goes a long way towards doing this.

Anyone with a bit of R skills now has easy functional access to property information for every location in the United States.

With a few lines of R code you will have access to at-least as much, and in most cases, MORE information than brokers.

You can take a look at a bit of what the package does in this introductory tutorial.

Functions

• geocode() : Batch geocoder
• listing_counts() : Area summary listings
• listings() : Area detailed listings
• vitality() : Area market vitality
• properties_near() : Listings near a location
• map_listings() : Featured listings from a map
• median_prices() : Area median prices
• parse_listing_urls() : Detailed property listing information
• mortgage_rates() : Rolling 30 day mortgage rates by product

Dictionaries

• dictionary_property_types() : Search-able property types
• dictionary_listing_features() : Search-able property features

Fun Tools

• summarise_broker_bullshit() : Summarizes a long-winded broker property description into n sentences using the PageRank algorithm

realtr's Issues

emo package during install

Going through the tutorial here:
http://asbcllc.com/r_packages/realtR/2018/introduction/

emo isn't installed through the normal installation process. I believe this is because it's github-only. Everything else seems to install (or I already had it) fine.

install.packages("devtools")

devtools::install_github("hadley/emo")

Tutorial Issues Object not found and repeated filter

Going through the tutorial here:
http://asbcllc.com/r_packages/realtR/2018/introduction/

df_alex_top <-
####THIS OBJECT DOES NOT EXIST
df_alex_school %>%

THIS LINE IS REPEATED TWICE

filter(scorePrediction == max(scorePrediction)) %>%
filter(scorePrediction == max(scorePrediction)) %>%
select(nameLocationSearch , scorePrediction:latitudeLocation) %>%
select(-one_of(c("idSchool", "codeCountry"))) %>%
purrr::set_names(c(
"search",
"score",
"school",
"address",
"city",
"zip",
"state",
"lat",
"lon"
))

map_listings() ignores homes with missing addresses

First off, as someone with a shared frustration of dealing with real-estate agents, thank you for creating this package.

As I was looking for properties using the search "02139, Cambridge, MA", I found that there were more properties listed directly on the realtor.com site (17 currently) than using the map_listings() function (14 currently). After some investigating, it looks like the 3 properties that aren't showing up are the only listings that don't list their exact address (despite being listed in the 02139 zip code search).

I imagine there are several homes listed on realtor.com that would be worth tracking via map_listings(), even without having an exact address. Not sure if you think it's possible or worth updating map_listings() to pull these listings as well, but thought it would be at least worthwhile to note in case you wanted to update documentation.

Apologies in advance if I am off-base here.

error on install

Hi,
Sounds like a great package. But when I try to install using devtools::install_github("abresler/realtR")

I get the following error:

Downloading GitHub repo abresler/realtR@master
from URL https://api.github.com/repos/abresler/realtR/zipball/master
Installing realtR
'/Library/Frameworks/R.framework/Resources/bin/R' --no-site-file --no-environ --no-save --no-restore --quiet CMD INSTALL
'/private/var/folders/r5/z80zbknj4dz_6pyxn_xqlfb80000gn/T/Rtmpy9OP12/devtools25412b1a4000/abresler-realtR-c802a00'
--library='/Library/Frameworks/R.framework/Versions/3.4/Resources/library' --install-tests

• installing source package ‘realtR’ ...
  ** R
  ** preparing package for lazy loading
  Error in loadNamespace(i, c(lib.loc, .libPaths()), versionCheck = vI[[i]]) :
  there is no package called ‘emo’
  ERROR: lazy loading failed for package ‘realtR’
• removing ‘/Library/Frameworks/R.framework/Versions/3.4/Resources/library/realtR’
  Installation failed: Command failed (1)

listings() function returns "No results"

Looks like realtor.com may have shutdown their API access for good? I'm not knowledgeable enough on this topic. Perhaps temporarily down for emergency maintenance.

https://www.realtor.com/robots.txt

realtor.com banning

I think there is an issue with realtor.com banning folks as this package scrapes the website. Any work arounds possible?

parse_listing_urls not working

I've used this exactly as you did in your tutorial, but it returns an empty tibble. It looks like it should accept listing API URL rather than plain listing URL. Perhaps they have changed the API? I also might have gotten a 403 error?

listings returns no result

Try to replicate the codes as I read in the Twitter, and it does not work.

library(realtR)
my_neighbors <- listings(locations = "Malibu, CA",
                                 beds_min = 4)

It gives error message:

Unknown variables: streetOrNeighborhoodSearch, zipcodeSearchUnknown variables: postal_codeJoining, by = "urlAPI"
Unknown variables: nameCity, stateSearch, zipcodeSearchJoining, by = "item"
Joining, by = "item"
Unknown variables: urlGeoAPI, urlAPI

Data Update

When using the trends_zipcode function the data returned is quite old (2018). Is this package pulling in revised market data? If not, is this something that can be added on a road map?

There are a lot of missing columns when using the listings function

I'm not sure if realtor.com changed the name's in the API, but I am getting a lot of unknown column messages.

Unknown columns: `statusListing`, `urlImage`, `addressProperty`, `cityProperty`, `stateProperty`, `zipcodeProperty`, `priceListing`, `areaPropertySF`, `countBaths`, `countBeds`, `sizeLotAcres`, `nameBrokerage`, `pricePerSFListing`