examples/DigitalDNA.md

In WAFI-CNR/ddna-rpackage: Digital DNA

DigitalDNA-Rpackage

WAFI-CNR 01/09/2019

Installation

install.packages("devtools", repos = "http://cran.us.r-project.org")
library(devtools)
devtools::install_github("WAFI-CNR/ddna-rpackage")

How to use

The “sequence_tweets_from_file” function call the Python function “digitaldna.TwitterDDNASequencer” and codifies json timeline according to three type of alphabets. Default alphabet “b3_type” adopts: (A) for tweets; (C) for replies; and (T) for retweets. More information on alphabets here: https://wafi-cnr.github.io/ddna/stable/generated/digitaldna.TwitterDDNASequencer.html

library(ddna)
df <- sequence_tweets_from_file("timelines.json", alphabet="b3_type")
str(df)

## 'data.frame':    6 obs. of  2 variables:
##  $ user_id: Factor w/ 6 levels "1135017996","22834067",..: 5 3 2 4 6 1
##  $ seq    : Factor w/ 6 levels "AAAAAAAAAAACAAACCAAAACAAAAAAAAAAAAAATAATAAATTAAAAAATCAAAACAAACAAAAAAACAAAACAACAAAACAACACAAAATACAAAACAAACCCCAACA"| __truncated__,..: 6 3 5 2 1 4

The package allows you to upload a CSV with already encoded DNA

df <- read.csv2("italian_retweets_users_sequences_new.csv", sep = ",")
str(df)

Plot Bases distribution

It is possible to observe the distribution of each base of the chosen alphabet. For each base it will be plotted a boxplot. The function accepts as a parameter the index of the dataframe column of the codified timeline (default 3).

plot_bases_distribution(df, dnacol = 3)

Plot interseq entropy

This function produces a composite plot. On the left a boxplot representing the distribution of the inter-sequence entropy (Shannon’s Entropy of the letters in the same with same sequence index but in different sequences). On the right a scatterplot of the entropies ordered by sequence index. “You can pass”plot_interseq" accept the column index in which there is the DNA sequence as a parameter (default=2).

plot_interseq(df, dnacol = 3)

Intrasequence Entropy plot

This function produces a box plot with a single box representing the distribution of the intra-sequence entropies (the Shannon Entropy computed over a single digital dna sequence).

plot_intraseq(df, dnacol = 3)

LCS plot

The lcs plot function compute the longest common sequences among DDNA strings. It accepts as parameters a threshold and a window. Threshold indicates possible bots accounts and it can be calculated on the lcs curve derived and smoothed by a running mean (threshold = “auto”) or assigned as a integer (threshold = x). Window parameter set the running mean window (neighbourhood). The lcs lenght column must be labelled as “dna”.

lcs_plot(df, threshold = "auto", window = 10)

Bots detection

“predicts_bots” produces a single column dataframe (logical) which identifies possible bot accounts.

bots = predict_bots(df, threshold = "auto", window = 10)

## bots
## FALSE  TRUE 
##   454   284

Plot Sequence Color

This function plots codified timelines colored according to the chosen alphabet. Timelines are plotted in deacresing order according to their length.

plot_sequences_by_color(df, dnacol = 3)

WAFI-CNR/ddna-rpackage documentation built on Oct. 31, 2019, 1:11 a.m.