def _quantile(self, valeur):
classe = self._classe_du_quantile(valeur)
a = classe.start
b = self.interval
c = _r(classe.getNext().getNext().frequence_cum -
classe.getNext().frequence_cum)
d = _r(classe.frequence_cum - classe.getPrev().frequence_cum)
print(f"{a} + {b} * ({c} / {d})\t\t{valeur}")
return a + b * (c / d)
def histogram(self):
histo = "\n\t\t\t c\t\tfreq\tfreq^\tfreq^%"
for i in self.classes:
interval = f"[{i.start}, {i.end}["
baton = "".join("#" for i in range(i.effectif))
centre = f"({_r(i.centre)})"
freq = _r(i.frequence)
freq_cum = _r(i.frequence_cum)
freq_cum_pc = _r(i.frequence_cum * 100)
histo += f"\n{interval}\t{centre}\t{freq} " \
f"\t{freq_cum} \t{freq_cum_pc} \t{baton} {len(baton)}"
return histo
def mode(self):
classe = self.classe_modale()[0]
a = classe.start
b = classe.effectif - classe.getPrev().effectif
d = classe.effectif - classe.getNext().effectif
c = self.interval
return _r(a + (b * c) / (b + d))
def _serialize_data(self):
"""Serialization des données"""
return {
'study': {
'general': {
'effectifs': self.effectifs,
'min': min(self.serie),
'max': max(self.serie),
'etendue': self.etendue,
'mode': self.mode(),
},
'centre': {
'centre': _r(self.centre),
'moyenne': _r(self.moyenne),
'mediane': _r(self.mediane()),
},
'dispersion': {
'ecart_absolu_moyen': _r(self.EAM()),
'ecart_semi_inter_quartile': _r(self.ESI()),
'variance': _r(self.variance()),
'ecart_type': _r(self.ecart_type()),
'coefficient_de_variation': {
'data': _r(self.coefficient_variation()),
'info': self._get_coeff_info('variation', COEF_VAR)
}
},
'forme': {
'coefficient_asymetrie': {
'data': _r(self.coeff_asym()),
'info': self._get_coeff_info('asymetrie')
},
'coefficient_applatissement': {
'data': _r(self.coeff_appl()),
'info': self._get_coeff_info('applatissement')
}
},
'quantiles': {
'quartiles': self.quartiles(),
'deciles': self.deciles(),
},
}
}
def getData(self):
data = self._serialize_data()
general = data['study']['general']
general['classe_modale'] = self.classe_modale()[0].range_repr
data['plot'] = {
"data": {i.range_repr: i.effectif
for i in self.classes},
# Delete ?
"classes": [i._serialize_data() for i in self.classes],
"freq": {
"ticks": [i.range_repr for i in self.classes],
"eff": [int(i.effectif) for i in self.classes],
"eff_pc": [_r(i.frequence_pc) for i in self.classes],
},
"cum": {
'ticks': [f"{i.start}{' ' * 5}{i.end}" for i in self.classes],
"freq": [_r(i.frequence_cum) for i in self.classes],
"eff": [int(i.effectif_cum) for i in self.classes]
}
}
return data
def _quantiles(self, type, population):
data = pd.DataFrame(population)
position = [_r((1 / type) * i) for i in range(1, type)]
quartiles = [_r(data.quantile(i)[0]) for i in position]
return {c + 1: i for c, i in enumerate(quartiles)}
def frequences(self, cumule=False, pourcent=False):
p = lambda: 100 if pourcent else 1
c = lambda x: x.frequence_cum if cumule else x.frequence
return [_r(c(i) * p()) for i in self.classes]
def quartiles(self):
q1 = self._quantile(0.25)
q2 = self.mediane()
q3 = self._quantile(0.75)
return {1: _r(q1), 2: _r(q2), 3: _r(q3)}