Skip to content

Introduction and dataset

After reading data into Raphtory we can now make use of the graph representation to ask some interesting questions. For this tutorial, we will use a dataset from SocioPatterns, comprising different behavioural interactions between a group of 22 baboons over a month.

Info

If you want to read more about the dataset, you can check it out in this paper: V. Gelardi, J. Godard, D. Paleressompoulle, N. Claidière, A. Barrat, “Measuring social networks in primates: wearable sensors vs. direct observations”, Proc. R. Soc. A 476:20190737 (2020).

In the below code we load this dataset into a dataframe and do a small amount of preprocessing to prepare it for loading into Raphtory. This includes dropping rows with blank fields and mapping the values of the behaviour category into a weight which can be aggregated. The mapping consists of the following conversions:

  • Affiliative (positive interaction) → +1
  • Agonistic (negative interaction) → -1
  • Other (neutral interaction) → 0

Graph 

import pandas as pd

edges_df = pd.read_csv(
    "data/OBS_data.txt", sep="\t", header=0, usecols=[0, 1, 2, 3, 4], parse_dates=[0]
)
edges_df["DateTime"] = pd.to_datetime(edges_df["DateTime"])
edges_df.dropna(axis=0, inplace=True)
edges_df["Weight"] = edges_df["Category"].apply(
    lambda c: 1 if (c == "Affiliative") else (-1 if (c == "Agonistic") else 0)
)
print(edges_df.head())

Output

              DateTime   Actor Recipient  Behavior     Category  Weight
15 2019-06-13 09:50:00  ANGELE    FELIPE  Grooming  Affiliative       1
17 2019-06-13 09:50:00  ANGELE    FELIPE  Grooming  Affiliative       1
19 2019-06-13 09:51:00  FELIPE    ANGELE   Resting  Affiliative       1
20 2019-06-13 09:51:00  FELIPE      LIPS   Resting  Affiliative       1
21 2019-06-13 09:51:00  ANGELE    FELIPE  Grooming  Affiliative       1

Next we load this into Raphtory using the load_edges_from_pandas function, modelling it as a weighted multi-layer graph, with a layer per unique behaviour.

Graph 

import raphtory as rp

g = rp.Graph()
g.load_edges_from_pandas(
    df=edges_df,
    src="Actor",
    dst="Recipient",
    time="DateTime",
    layer_col="Behavior",
    properties=["Weight"],
)
print(g)

Output

Graph(number_of_nodes=22, number_of_edges=290, number_of_temporal_edges=3196, earliest_time=1560419400000, latest_time=1562756700000)