Some of the code behind @petebutbot

petebutbot.py

# I scraped exisisting tweets using R 

# library(rtweet)
# library(tidyverse)

# petey <- get_timeline("PeteButtigieg", n = 3200, include_rts = FALSE, exclude_replies = FALSE)

# petey %>%
# select(text) %>%
# rename(tweets = text) %>%
# write_csv("petebuttigieg_tweets.csv")


import markovify
# Get raw text as string.
with open("petebuttigieg_tweets.csv") as f:
    text = f.read()

# Build the model.
text_model = markovify.Text(text)

# # Print five randomly-generated sentences
# for i in range(5):
#     print(text_model.make_sentence())

# Print three randomly-generated sentences of no more than 280 characters
jsonfile = {"origin" : []}
stopWords = ["@", "https://"]
while len(jsonfile["origin"]) <= 365:
    tweet = text_model.make_short_sentence(280)
    shouldInclude = all([stopword not in tweet for stopword in stopWords])
    if shouldInclude:
        jsonfile["origin"].append(tweet)
print('{ \n "origin": \n[')
[print('"' + elem + '", \n') for elem in jsonfile["origin"]]
print("]}")