VI. Geography¶
In this section, we make use of the Geographic Data for Ancient Near Eastern Archaeological Sites from Professor Olof Pedersén at Uppsala University to locate some of the geographic names appeared in the Drehem texts.
We will use fuzzy search to find matches between the ANE files and our texts, and then use the geographic locations in the ANE to plot those that appear in our texts.
The following installs and imports the relevant libararies.
from google.colab import drive
drive.mount('/content/drive')
Drive already mounted at /content/drive; to attempt to forcibly remount, call drive.mount("/content/drive", force_remount=True).
%cd "drive/My Drive/SumerianNetworks"
[Errno 2] No such file or directory: 'drive/My Drive/SumerianNetworks'
/content/drive/My Drive/SumerianNetworks
!pip install geopandas
!pip install descartes
!pip install pysal
!pip install fuzzywuzzy
!pip install python-Levenshtein
!pip install NetworkX
import pandas as pd
import numpy as np
import geopandas as gpd
from fuzzywuzzy import fuzz
from fuzzywuzzy import process
import fiona
fiona.drvsupport.supported_drivers['kml'] = 'rw' # enable KML support which is disabled by default
fiona.drvsupport.supported_drivers['KML'] = 'rw' # enable KML support which is disabled by defaults
import networkx as nx
/usr/local/lib/python3.7/dist-packages/geopandas/_compat.py:110: UserWarning: The Shapely GEOS version (3.8.0-CAPI-1.13.1 ) is incompatible with the GEOS version PyGEOS was compiled with (3.9.1-CAPI-1.14.2). Conversions between both will be slow.
shapely_geos_version, geos_capi_version_string
We first import filtered Drehem data and extract all the geographic names: SN (settlement name), TN (temple name), AN, FN (field name), QN (quarter name), WN (water course name).
drehem = pd.read_csv("JupyterBook/words_df.csv")
drehem.head()
| Unnamed: 0 | lemma | id_text | id_line | id_word | label | |
|---|---|---|---|---|---|---|
| 0 | 0 | 1(geš₂)[]NU | P113959 | 3 | P113959.3.1 | o 1 |
| 1 | 1 | 1(u)[]NU | P113959 | 3 | P113959.3.2 | o 1 |
| 2 | 2 | 6(aš)[]NU | P113959 | 3 | P113959.3.3 | o 1 |
| 3 | 3 | 4(barig)[]NU | P113959 | 3 | P113959.3.4 | o 1 |
| 4 | 4 | 3(ban₂)[]NU | P113959 | 3 | P113959.3.5 | o 1 |
drehem.shape
(3175331, 6)
geo_names = {}
p_number = ''
places = []
geo_names_list = []
#SN (settlement name), TN (temple name), AN, FN (field name), QN (quarter name), WN (water course name)
#remove year names
types = ["SN", "TN", "AN", "FN", "QN", "WN"]
for index, row in drehem.iterrows():
# if row['id_word'][:7] != p_number:
# geo_names[p_number] = places
# p_number = row['id_word'][:7]
# places = []
w = row['lemma']
if w[-2:] in types:
geo_names[row['id_text']] = w
# geo_names[p_number] = [places]
# del geo_names['']
d = {'ip':list(geo_names.keys()),'geo_name': list(geo_names.values())}
geo_df = pd.DataFrame.from_dict(data = d)
geo_df.head()
| ip | geo_name | |
|---|---|---|
| 0 | P125538 | Nibru[1]SN |
| 1 | P379198 | hub₂-e[]FN |
| 2 | P110676 | e-su-dar[]FN |
| 3 | P388027 | Irisaŋrig[1]SN |
| 4 | P118246 | Gaeš[1]SN |
Now we remove the suffixes and leave just the geographical names. (eg. “[00]SN”).
def remove_tail(string):
return string.split("[")[0]
non_emp_geo = geo_df
non_emp_geo["cleaned_geo_name"] = non_emp_geo["geo_name"].map(remove_tail)
non_emp_geo.head()
| ip | geo_name | cleaned_geo_name | |
|---|---|---|---|
| 0 | P125538 | Nibru[1]SN | Nibru |
| 1 | P379198 | hub₂-e[]FN | hub₂-e |
| 2 | P110676 | e-su-dar[]FN | e-su-dar |
| 3 | P388027 | Irisaŋrig[1]SN | Irisaŋrig |
| 4 | P118246 | Gaeš[1]SN | Gaeš |
We now import the ANE KMZ files. Make sure you have the latest files from here and change the file name below accordingly.
location = gpd.read_file('JupyterBook/jupyter_collection/04_Geographic_Names/geopandas_intro/data/sites.kml', driver='KML',layer=2)
print(location.shape)
location.head()
(2857, 3)
| Name | Description | geometry | |
|---|---|---|---|
| 0 | Adab (Bismaya) | POINT Z (45.62386 31.95071 0.00000) | |
| 1 | Abydos | POINT Z (31.91906 26.18509 0.00000) | |
| 2 | Adamdun? (Teppe Surkhehgan) | POINT Z (48.79814 32.02168 0.00000) | |
| 3 | Admannu / Natmane (Tell 'Ali) | POINT Z (43.68200 35.38198 0.00000) | |
| 4 | Adumatu? (Dumat el-Jandal) | POINT Z (39.86723 29.81135 0.00000) |
Since the Name column contains both the modern name and ancient name of the same place, we will convert them into a list for matching purpose.
def parse_names(string):
"""Takes in a string (eg. Adamdun? (Teppe Surkhehgan))
Returns a tuple of list of names, and whether
or not there's doubts ('?') on the ancient name (0/1).
In case there
"""
names = []
ancient_names = ""
modern_names = ""
if '?' in string:
string = string.replace('?', '')
if '(' in string:
ancient_names = string.split('(')[0]
modern_names = string.split('(')[1].split(')')[0]
if '/' in modern_names:
for name in modern_names.split('/'):
names.append(name)
else:
names.append(modern_names)
else:
ancient_names = string
if '/' in ancient_names:
for name in ancient_names.split('/'):
names.append(name)
else:
names.append(ancient_names)
return names
location["Name (in list)"] = location["Name"].map(parse_names)
location.head()
| Name | Description | geometry | Name (in list) | |
|---|---|---|---|---|
| 0 | Adab (Bismaya) | POINT Z (45.62386 31.95071 0.00000) | [Bismaya, Adab ] | |
| 1 | Abydos | POINT Z (31.91906 26.18509 0.00000) | [Abydos] | |
| 2 | Adamdun? (Teppe Surkhehgan) | POINT Z (48.79814 32.02168 0.00000) | [Teppe Surkhehgan, Adamdun ] | |
| 3 | Admannu / Natmane (Tell 'Ali) | POINT Z (43.68200 35.38198 0.00000) | [Tell 'Ali, Admannu , Natmane ] | |
| 4 | Adumatu? (Dumat el-Jandal) | POINT Z (39.86723 29.81135 0.00000) | [Dumat el-Jandal, Adumatu ] |
There are also columns with a question mark indicating if there’s any doubt in the ancient and moden name equivalence, we will make a new column Doubt? indicating that information.
location["Doubt?"] = location["Name"].map(lambda string: '?' in string)
location.head()
| Name | Description | geometry | Name (in list) | Doubt? | |
|---|---|---|---|---|---|
| 0 | Adab (Bismaya) | POINT Z (45.62386 31.95071 0.00000) | [Bismaya, Adab ] | False | |
| 1 | Abydos | POINT Z (31.91906 26.18509 0.00000) | [Abydos] | False | |
| 2 | Adamdun? (Teppe Surkhehgan) | POINT Z (48.79814 32.02168 0.00000) | [Teppe Surkhehgan, Adamdun ] | True | |
| 3 | Admannu / Natmane (Tell 'Ali) | POINT Z (43.68200 35.38198 0.00000) | [Tell 'Ali, Admannu , Natmane ] | False | |
| 4 | Adumatu? (Dumat el-Jandal) | POINT Z (39.86723 29.81135 0.00000) | [Dumat el-Jandal, Adumatu ] | True |
Now expand location into one name for each row.
location_expanded = location.explode("Name (in list)")
location_expanded.head()
| Name | Description | geometry | Name (in list) | Doubt? | |
|---|---|---|---|---|---|
| 0 | Adab (Bismaya) | POINT Z (45.62386 31.95071 0.00000) | Bismaya | False | |
| 0 | Adab (Bismaya) | POINT Z (45.62386 31.95071 0.00000) | Adab | False | |
| 1 | Abydos | POINT Z (31.91906 26.18509 0.00000) | Abydos | False | |
| 2 | Adamdun? (Teppe Surkhehgan) | POINT Z (48.79814 32.02168 0.00000) | Teppe Surkhehgan | True | |
| 2 | Adamdun? (Teppe Surkhehgan) | POINT Z (48.79814 32.02168 0.00000) | Adamdun | True |
Now we have 5856 places.
location_expanded.shape
(5856, 5)
Now we have both files (ANE file and our Drehem file) ready, we will match the geo names in these files. We will use the fuzzywuzzy library’s WRatio, which is more tolerant to small differences like capitalization and punctuation. You can try different methods.
Note that the matching takes a fairly long time (> 30 mins) to run.
def fuzzy_search(geo_name):
scores = location_expanded['Name (in list)'].map(lambda name: fuzz.WRatio(geo_name, name))
return list(location_expanded[scores>90]['Name'])
non_emp_geo['fuzzy_match'] = non_emp_geo['cleaned_geo_name'].map(fuzzy_search)
non_emp_geo[non_emp_geo['fuzzy_match'].map(len)>0]
| ip | geo_name | cleaned_geo_name | fuzzy_match | |
|---|---|---|---|---|
| 9 | P121982 | Umma[1]SN | Umma | [Umma (Tell Jokha)] |
| 16 | P465780 | Umma[1]SN | Umma | [Umma (Tell Jokha)] |
| 19 | P141679 | Umma[1]SN | Umma | [Umma (Tell Jokha)] |
| 27 | P100676 | Umma[1]SN | Umma | [Umma (Tell Jokha)] |
| 56 | P131594 | Umma[1]SN | Umma | [Umma (Tell Jokha)] |
| ... | ... | ... | ... | ... |
| 20679 | P136415 | Lagaš[1]SN | Lagaš | [Lagaš (El-Hiba)] |
| 20684 | P122006 | Umma[1]SN | Umma | [Umma (Tell Jokha)] |
| 20686 | P218142 | Umma[1]SN | Umma | [Umma (Tell Jokha)] |
| 20690 | P322761 | Zabalam[1]SN | Zabalam | [Zabalam (Tulul Ibzaikh)] |
| 20697 | P378821 | Isin[1]SN | Isin | [Isin (Ishan Bahriyat)] |
3802 rows × 4 columns
If we set the matching score threshold to 90, we got 37 unique matches.
matches = non_emp_geo[non_emp_geo['fuzzy_match'].map(len)>0]
unique_matches = matches.drop_duplicates('geo_name').reset_index(drop=True)
# unique_matches = matched['cleaned_geo_name'].unique()
unique_matches.head()
| ip | geo_name | cleaned_geo_name | fuzzy_match | |
|---|---|---|---|---|
| 0 | P121982 | Umma[1]SN | Umma | [Umma (Tell Jokha)] |
| 1 | P136601 | Isin[1]SN | Isin | [Isin (Ishan Bahriyat)] |
| 2 | P110363 | Anšan[1]SN | Anšan | [Anšan (Tell Malyan)] |
| 3 | P324788 | Zabalam[1]SN | Zabalam | [Zabalam (Tulul Ibzaikh)] |
| 4 | P118180 | Eridug[1]SN | Eridug | [Eridu (Tell Abu Shahrain)] |
unique_matches['fuzzy_match'] = unique_matches['fuzzy_match'].map(lambda l: l[0])
unique_matches.head()
| ip | geo_name | cleaned_geo_name | fuzzy_match | |
|---|---|---|---|---|
| 0 | P121982 | Umma[1]SN | Umma | Umma (Tell Jokha) |
| 1 | P136601 | Isin[1]SN | Isin | Isin (Ishan Bahriyat) |
| 2 | P110363 | Anšan[1]SN | Anšan | Anšan (Tell Malyan) |
| 3 | P324788 | Zabalam[1]SN | Zabalam | Zabalam (Tulul Ibzaikh) |
| 4 | P118180 | Eridug[1]SN | Eridug | Eridu (Tell Abu Shahrain) |
Here we saved the matches to drive so we can reuse it later.
non_emp_geo.to_csv("JupyterBook/jupyter_collection/04_Geographic_Names/working files/non_emp_geo.csv")
Now we try to set the score to be > 87.5. Note the following cell also runs a fairly long time.
def fuzzy_search_875(geo_name):
scores = location_expanded['Name (in list)'].map(lambda name: fuzz.WRatio(geo_name, name))
return list(location_expanded[scores>87.5]['Name'])
non_emp_geo = pd.read_csv("JupyterBook/jupyter_collection/04_Geographic_Names/working files/non_emp_geo.csv")
non_emp_geo['fuzzy_match_875'] = non_emp_geo['cleaned_geo_name'].map(fuzzy_search_875)
non_emp_geo[non_emp_geo['fuzzy_match_875'].map(len)>0]
We got more matches, but some of them are not what we want.
non_emp_geo[non_emp_geo['fuzzy_match_875'].map(len)>0]["cleaned_geo_name"].unique()
We can see the subset of places which are not in the 33 matches with scores greater than 90 but are in matches with scores greater than 87.5. We can see some of them are not the same place, but there are places we missed in the 33.
in875_notin_90 = non_emp_geo[(non_emp_geo['fuzzy_match'].map(len)==2)&(non_emp_geo['fuzzy_match_875'].map(len)>0)]
in875_notin_90.head(50)
in875_notin_90[in875_notin_90["fuzzy_match_875"].map(lambda lst: any([s.find('Adamdun') != -1 for s in lst]))]
Since manually going over the names takes a lot of time and labor, we will focus on the 33 good matches at this time.
Here we will add the coordinate information to the matched locations.
def add_coordinates(str):
s = location[location["Name"]==str]['geometry']
if len(s) > 0:
return s.iloc[0]
else:
return None
unique_matches['coordinates'] = unique_matches["fuzzy_match"].map(add_coordinates)
unique_matches = unique_matches.dropna().reset_index(drop=True)
unique_matches.head()
| ip | geo_name | cleaned_geo_name | fuzzy_match | coordinates | |
|---|---|---|---|---|---|
| 0 | P121982 | Umma[1]SN | Umma | Umma (Tell Jokha) | POINT Z (45.88767609164621 31.66743001625768 0) |
| 1 | P136601 | Isin[1]SN | Isin | Isin (Ishan Bahriyat) | POINT Z (45.26937063450065 31.88437701508951 0) |
| 2 | P110363 | Anšan[1]SN | Anšan | Anšan (Tell Malyan) | POINT Z (52.41066812180781 30.01111323889061 0) |
| 3 | P324788 | Zabalam[1]SN | Zabalam | Zabalam (Tulul Ibzaikh) | POINT Z (45.87575461093311 31.74472924832169 0) |
| 4 | P118180 | Eridug[1]SN | Eridug | Eridu (Tell Abu Shahrain) | POINT Z (45.99672447851623 30.81686971277664 0) |
Now we will use NetworkX to plot the locations in a graph. We have a few additional libraries to install and import.
!pip install contextily
!pip install cartopy
!pip uninstall shapely
!pip install shapely --no-binary shapely
Requirement already satisfied: contextily in /usr/local/lib/python3.7/dist-packages (1.1.0)
Requirement already satisfied: joblib in /usr/local/lib/python3.7/dist-packages (from contextily) (1.0.1)
Requirement already satisfied: requests in /usr/local/lib/python3.7/dist-packages (from contextily) (2.23.0)
Requirement already satisfied: pillow in /usr/local/lib/python3.7/dist-packages (from contextily) (7.1.2)
Requirement already satisfied: matplotlib in /usr/local/lib/python3.7/dist-packages (from contextily) (3.2.2)
Requirement already satisfied: mercantile in /usr/local/lib/python3.7/dist-packages (from contextily) (1.2.1)
Requirement already satisfied: rasterio in /usr/local/lib/python3.7/dist-packages (from contextily) (1.2.3)
Requirement already satisfied: geopy in /usr/local/lib/python3.7/dist-packages (from contextily) (1.17.0)
Requirement already satisfied: certifi>=2017.4.17 in /usr/local/lib/python3.7/dist-packages (from requests->contextily) (2020.12.5)
Requirement already satisfied: chardet<4,>=3.0.2 in /usr/local/lib/python3.7/dist-packages (from requests->contextily) (3.0.4)
Requirement already satisfied: urllib3!=1.25.0,!=1.25.1,<1.26,>=1.21.1 in /usr/local/lib/python3.7/dist-packages (from requests->contextily) (1.25.11)
Requirement already satisfied: idna<3,>=2.5 in /usr/local/lib/python3.7/dist-packages (from requests->contextily) (2.10)
Requirement already satisfied: numpy>=1.11 in /usr/local/lib/python3.7/dist-packages (from matplotlib->contextily) (1.19.5)
Requirement already satisfied: pyparsing!=2.0.4,!=2.1.2,!=2.1.6,>=2.0.1 in /usr/local/lib/python3.7/dist-packages (from matplotlib->contextily) (2.4.7)
Requirement already satisfied: kiwisolver>=1.0.1 in /usr/local/lib/python3.7/dist-packages (from matplotlib->contextily) (1.3.1)
Requirement already satisfied: python-dateutil>=2.1 in /usr/local/lib/python3.7/dist-packages (from matplotlib->contextily) (2.8.0)
Requirement already satisfied: cycler>=0.10 in /usr/local/lib/python3.7/dist-packages (from matplotlib->contextily) (0.10.0)
Requirement already satisfied: click>=3.0 in /usr/local/lib/python3.7/dist-packages (from mercantile->contextily) (7.1.2)
Requirement already satisfied: cligj>=0.5 in /usr/local/lib/python3.7/dist-packages (from rasterio->contextily) (0.7.1)
Requirement already satisfied: affine in /usr/local/lib/python3.7/dist-packages (from rasterio->contextily) (2.3.0)
Requirement already satisfied: snuggs>=1.4.1 in /usr/local/lib/python3.7/dist-packages (from rasterio->contextily) (1.4.7)
Requirement already satisfied: click-plugins in /usr/local/lib/python3.7/dist-packages (from rasterio->contextily) (1.1.1)
Requirement already satisfied: attrs in /usr/local/lib/python3.7/dist-packages (from rasterio->contextily) (21.2.0)
Requirement already satisfied: geographiclib<2,>=1.49 in /usr/local/lib/python3.7/dist-packages (from geopy->contextily) (1.50)
Requirement already satisfied: six>=1.5 in /usr/local/lib/python3.7/dist-packages (from python-dateutil>=2.1->matplotlib->contextily) (1.15.0)
Requirement already satisfied: cartopy in /usr/local/lib/python3.7/dist-packages (0.19.0.post1)
Requirement already satisfied: pyshp>=2 in /usr/local/lib/python3.7/dist-packages (from cartopy) (2.1.3)
Requirement already satisfied: shapely>=1.5.6 in /usr/local/lib/python3.7/dist-packages (from cartopy) (1.7.1)
Requirement already satisfied: numpy>=1.13.3 in /usr/local/lib/python3.7/dist-packages (from cartopy) (1.19.5)
Uninstalling Shapely-1.7.1:
Would remove:
/usr/local/lib/python3.7/dist-packages/Shapely-1.7.1-py3.7.egg-info
/usr/local/lib/python3.7/dist-packages/shapely/*
Proceed (y/n)? y
Successfully uninstalled Shapely-1.7.1
Collecting shapely
Using cached https://files.pythonhosted.org/packages/42/f3/0e1bc2c4f15e05e30c6b99322b9ddaa2babb3f43bc7df2698efdc1553439/Shapely-1.7.1.tar.gz
Skipping wheel build for shapely, due to binaries being disabled for it.
ERROR: albumentations 0.1.12 has requirement imgaug<0.2.7,>=0.2.5, but you'll have imgaug 0.2.9 which is incompatible.
Installing collected packages: shapely
Running setup.py install for shapely ... ?25l?25hdone
Successfully installed shapely-1.7.1
from libpysal import weights, examples
from contextily import add_basemap
import matplotlib.pyplot as plt
import networkx as nx
import cartopy.crs as ccrs
import matplotlib.pyplot as plt
import cartopy
/usr/local/lib/python3.7/dist-packages/geopandas/_compat.py:110: UserWarning: The Shapely GEOS version (3.6.2-CAPI-1.10.2 4d2925d6) is incompatible with the GEOS version PyGEOS was compiled with (3.9.1-CAPI-1.14.2). Conversions between both will be slow.
shapely_geos_version, geos_capi_version_string
GeoPandas dataframe requires a geometry attribute, we construct geo_um.
coordinates = np.column_stack((unique_matches["coordinates"].map(lambda p: p.x), unique_matches["coordinates"].map(lambda p: p.y)))
geo_um = gpd.GeoDataFrame(unique_matches, geometry = list(unique_matches['coordinates']))
geo_um.head()
| ip | geo_name | cleaned_geo_name | fuzzy_match | coordinates | geometry | |
|---|---|---|---|---|---|---|
| 0 | P121982 | Umma[1]SN | Umma | Umma (Tell Jokha) | POINT Z (45.88767609164621 31.66743001625768 0) | POINT Z (45.88768 31.66743 0.00000) |
| 1 | P136601 | Isin[1]SN | Isin | Isin (Ishan Bahriyat) | POINT Z (45.26937063450065 31.88437701508951 0) | POINT Z (45.26937 31.88438 0.00000) |
| 2 | P110363 | Anšan[1]SN | Anšan | Anšan (Tell Malyan) | POINT Z (52.41066812180781 30.01111323889061 0) | POINT Z (52.41067 30.01111 0.00000) |
| 3 | P324788 | Zabalam[1]SN | Zabalam | Zabalam (Tulul Ibzaikh) | POINT Z (45.87575461093311 31.74472924832169 0) | POINT Z (45.87575 31.74473 0.00000) |
| 4 | P118180 | Eridug[1]SN | Eridug | Eridu (Tell Abu Shahrain) | POINT Z (45.99672447851623 30.81686971277664 0) | POINT Z (45.99672 30.81687 0.00000) |
We try to build connections using p-numbers: since the same geographic name can appear in multiple p-numbers, so we add an edge if two places appear in the same p-number. But unfortunately, we can’t find two places appearing in the same p-number, so instead we plotted a fully connected graph.
matches.head()
| ip | geo_name | cleaned_geo_name | fuzzy_match | |
|---|---|---|---|---|
| 9 | P121982 | Umma[1]SN | Umma | [Umma (Tell Jokha)] |
| 16 | P465780 | Umma[1]SN | Umma | [Umma (Tell Jokha)] |
| 19 | P141679 | Umma[1]SN | Umma | [Umma (Tell Jokha)] |
| 27 | P100676 | Umma[1]SN | Umma | [Umma (Tell Jokha)] |
| 56 | P131594 | Umma[1]SN | Umma | [Umma (Tell Jokha)] |
name_pnum = matches.groupby(['cleaned_geo_name'])['ip'].apply(list).reset_index(name = 'pnum')
name_pnum.head()
| cleaned_geo_name | pnum | |
|---|---|---|
| 0 | Adab | [P325875, P126986, P458586, P123317, P201858, ... |
| 1 | Anšan | [P110363, P356008, P113518, P406257, P131214, ... |
| 2 | Awal | [P107255, P145800, P126531, P107256, P112023] |
| 3 | Aššur | [P117479, P143751, P134727, P248907, P126176] |
| 4 | Babila | [P101457, P129473, P119451, P103818] |
g = nx.Graph()
#Add nodes to the graph
for place in name_pnum['cleaned_geo_name']:
g.add_node(place)
#Add edges
num_edges = 0
for place1 in name_pnum['cleaned_geo_name'].values:
for place2 in name_pnum['cleaned_geo_name'].values:
if place1 != place2:
weight = 1
# for pnum1 in name_pnum[name_pnum['cleaned_geo_name']==place1]['pnum'].values[0]:
# for pnum2 in name_pnum[name_pnum['cleaned_geo_name']==place2]['pnum'].values[0]:
# if pnum1 == pnum2:
# weight += 1
# num_edges += 1
g.add_edge(place1, place2)
pos = {}
for place in name_pnum['cleaned_geo_name'].values:
loc = unique_matches[unique_matches['cleaned_geo_name']==place]["coordinates"].values[0]
pos[place] = (loc.x, loc.y)
crs = ccrs.PlateCarree()
fig, ax = plt.subplots(
1, 1, figsize=(12, 8), subplot_kw=dict(projection=crs))
ax.add_feature(cartopy.feature.OCEAN)
ax.add_feature(cartopy.feature.LAND, edgecolor='black')
ax.add_feature(cartopy.feature.LAKES, edgecolor='black')
ax.add_feature(cartopy.feature.RIVERS)
ax.gridlines()
# Extent of .
ax.set_extent([30, 55, 25, 40])
nx.draw_networkx(g, ax=ax,
font_size=10,
alpha=.8,
width=.075,
pos=pos,
cmap=plt.cm.autumn)
We can zoom in a little to the area where we have a lot of known geo names (around Drehem).
crs = ccrs.PlateCarree()
fig, ax = plt.subplots(
1, 1, figsize=(12, 8), subplot_kw=dict(projection=crs))
ax.add_feature(cartopy.feature.OCEAN)
ax.add_feature(cartopy.feature.LAND, edgecolor='black')
ax.add_feature(cartopy.feature.LAKES, edgecolor='black')
ax.add_feature(cartopy.feature.RIVERS)
ax.gridlines()
# Extent of .
ax.set_extent([42, 48, 30, 35])
nx.draw_networkx(g, ax=ax,
font_size=10,
alpha=.8,
width=.075,
pos=pos,
cmap=plt.cm.autumn)
This is in no way completed, potential next step might include interactive graphs like those realized in Gephi (potentially using Bokeh?) and explore other methods to set up edges (currently there are no two places with a same p number so I just simply connected them) and edge weights.