analyse_twit

analyse.py

from datetime import datetime
from math import sqrt
from json import load as _json_load
import numpy as np
from scipy import stats # for regression line
import pylab


def get_datetime(s_date):
    year, month ,day = s_date.split('/')
    year = int(year)
    month = int(month)
    day = int(day)
    return datetime(year, month ,day)

def get_numbers(twits):
    datetime_twits = {}
    for date, number in twits.items():
        datetime_twits[get_datetime(date)] = number
    twits = datetime_twits

    numbers = []
    for date in sorted(twits.keys()):
        numbers.append(twits[date])
    return numbers


def parser(f):
    twits = _json_load(f)
    return get_numbers(twits)

def statics_repr(numbers):
    sloop, intercept, *_ = stats.linregress(range(len(numbers)), numbers)
    s = 'average: {}\n'.format(np.average(numbers))
    s += 'variance: {}\n'.format(sqrt(np.var(numbers)))
    s += 'regression line: {} * x + {}\n'.format(sloop, intercept)
    s += 'correlation coefficient: {}\n'.format(np.corrcoef(numbers,
        range(len(numbers)))[0][1])
    return s

def plot(numbers):
    # plot from origin
    origin = pylab.arange(1), pylab.arange(1)
    pylab.plot(*origin)

    # plot line graph
    x = pylab.arange(0, len(numbers))
    y = numbers
    pylab.plot(x, y)

    # plot regression line
    sloop, intercept, *_ = stats.linregress(range(len(numbers)), numbers)
    x = pylab.arange(0, len(numbers))
    y = sloop * x + intercept
    pylab.plot(x,y)

    pylab.show()

if __name__ == '__main__':
    from sys import argv
    filename = argv[1]
    numbers = parser(open(filename))
    print(statics_repr(numbers))
    plot(numbers)