Merge pull request #13 from UPB-SS1/crowdtangle_optional_parameters

Crowdtangle optional parameters

README.md

@@ -14,6 +14,9 @@ CLSB refers to a specific coordinated activity performed by a network of Faceboo
 To identify such networks, we designed, implemented and tested an algorithm that detects sets of Facebook public entities which performed CLSB by (1) estimating a time threshold that identifies URLs shares performed by multiple distinguished entities within an unusually short period of time (as compared to the entire dataset), and (2) grouping the entities that repeatedly shared the same news story within this coordination interval. The rationale is that, while it may be common that several entities share the same URLs, it is unlikely, unless a consistent coordination exists, that this occurs within the time threshold and repeatedly.
 
 ## Installation
+
+[https://pypi.org/project/pycoornet/](https://pypi.org/project/pycoornet/)
+
 ```sh
 pip install pycoornet
 ```
@@ -36,7 +39,7 @@ def main():
     links_df = pd.read_csv('samples/sample_source_links.csv')
     # Init CrowdTangle with api key
     crowd_tangle = CrowdTangle("abc123def345")
-    ct_df = crowd_tangle.get_shares(urls=links_df, url_column='clean_url', date_column='date',clean_urls=True, platforms='facebook', sleep_time=1)
+    ct_df = crowd_tangle.get_shares(urls=links_df, url_column='clean_url', date_column='date',clean_urls=True, platforms='facebook', sleep_time=30)
     shared = Shared()
     crowtangle_shares_df, shares_graph, q = shared.coord_shares(ct_df, clean_urls=True)
 

setup.py

@@ -14,12 +14,13 @@
 
 setup(
     name="pycoornet",
-    version="0.2.2",
+    version="0.2.3",
     description="Using Python Given a set of URLs, this packages detects coordinated link sharing behavior on social media and outputs the network of entities that performed such behaviour.",
     long_description=long_description,
     long_description_content_type="text/markdown",
     author = 'Camilo Andres Soto Montoya, Jose R. Zapata',
-    author_email = '[email protected], ',
+    author_email = '[email protected], [email protected]',
+
     url="https://github.com/UPB-SS1/PyCooRnet",
     packages=find_packages(where='src' ,exclude=["tests", "*.test", "*.tes.*"]),
     package_dir={

src/pycoornet/crowdtangle.py

@@ -24,7 +24,8 @@ def __init__(self, api_key):
             raise Exception('Crowdtangle Api Token is missing')
         self.api_key = api_key
 
-    def get_shares(self, urls, url_column='url', date_column='date', platforms=('facebook', 'instagram'), nmax=500, sleep_time=20, clean_urls=False, save_ctapi_output=False):
+    def get_shares(self, urls, url_column='url', date_column='date', platforms=('facebook', 'instagram'),
+                   nmax=500, sleep_time=20, clean_urls=False, save_ctapi_output=False, id_column=None, remove_days=None):
         """ Get the URLs shares from CrowdTangle from a list of URLs with publish datetime
 
         Args:
@@ -38,7 +39,8 @@ def get_shares(self, urls, url_column='url', date_column='date', platforms=('fac
                                         depending on the assigned API rate limit. Defaults to 20.
             clean_urls (bool, optional): clean the URLs from tracking parameters. Defaults to False.
             save_ctapi_output (bool, optional): saves the original CT API output in rawdata/ folder. Defaults to False.
-
+            id_column(str,optional): name of the column wherre the id of each URL is stored.
+            remove_days(int,optional): remove shares performed more than X days from first share
         Raises:
             Exception: [description]
             Exception: [description]
@@ -146,8 +148,16 @@ def get_shares(self, urls, url_column='url', date_column='date', platforms=('fac
                     df_full['date'] = pd.to_datetime(df_full['date'])
                     df_full = df_full.set_index('date', drop=False)
 
+                    # if id column is specified
+                    if id_column:
+                        df_full["id_column"] = urls.iloc[i, :].loc[id_column]
+
+
                     # remove shares performed more than one week from first share
-                    df_full = df_full.loc[(df_full.index <= df_full.index.min()+ pd.Timedelta('7 day'))]
+                    if remove_days:
+                        # ex: '7 day'
+                        days = f"{remove_days} day"
+                        df_full = df_full.loc[(df_full.index <= df_full.index.min()+ pd.Timedelta(days))]
 
                     # concat data results in dataframe
                     ct_shares_df = ct_shares_df.append(df_full, ignore_index=True)

tests/test_pycoornet.py

@@ -17,8 +17,10 @@ def sample_source_df():
 def test_crowdtangle(crowd_token, sample_source_df):
     crowd_tangle = CrowdTangle(crowd_token)
     shares_df = crowd_tangle.get_shares(urls=sample_source_df, url_column='clean_url', date_column='date',
-                                        clean_urls=True, platforms='facebook', sleep_time=1)
+                                        clean_urls=True, platforms='facebook', sleep_time=1,
+                                        id_column = 'url_rid')
     if shares_df.shape[0] > 0:
+
         assert True
     else:
         assert False