student_analysis.py

import numpy as np

#~ TODO:
#~ FAZER CACHE DE TUDO
#~ AO CHAMAR A FUNCAO VERIFICAR SE TEM ALGO NA CACHE

from script.utils.situations import *
import pandas as pd

CURRENT_YEAR = 2017
CURRENT_SEMESTER = 1


def list_students(df):
	#~ ativos = df[["MATR_ALUNO", "NOME_PESSOA",]][df["FORMA_EVASAO"] == EvasionForm.EF_ATIVO].drop_duplicates()
	situations = df.groupby(["MATR_ALUNO", "NOME_PESSOA", "FORMA_EVASAO"])
	situations = list(pd.DataFrame({'count' : situations.size()}).reset_index().groupby(["FORMA_EVASAO"]))
	#~ Cria lista de nome de listagens
	list_situations = {}
	for s in situations:
		#Busca a lista de alunos relacionados a um codigo
		list_situations[s[0]] = list(s[1]["MATR_ALUNO"])

	return list_situations

def ira_alunos(df):
	iras = ira_por_quantidade_disciplinas(df)
	for i in iras:
		ira_total = 0
		carga_total = 0
		for semestre in iras[i]:
			ira_total += iras[i][semestre][0]*iras[i][semestre][2]
			carga_total+=iras[i][semestre][2]
		
		if(carga_total != 0):
			iras[i] = ira_total/carga_total
		else:
			iras[i] = 0
	return iras
	
def taxa_aprovacao(df):
	aprovacoes_semestres = indice_aprovacao_semestral(df)
	
	for aluno in aprovacoes_semestres:
		total = sum([aprovacoes_semestres[aluno][s][1] for s in aprovacoes_semestres[aluno]])
		aprovacoes = sum([aprovacoes_semestres[aluno][s][0] for s in aprovacoes_semestres[aluno]])
		total = float(total)
		aprovacoes = float(aprovacoes)
		if(total != 0):
			aprovacoes_semestres[aluno] = aprovacoes/total
		else:
			aprovacoes_semestres[aluno] = None
		#~ for semestre in aprovacoes_semestres[aluno]:
			#~ aprovacoes+=aprovacoes_semestres[aluno][semestre][0]
			#~ total+=aprovacoes_semestres[semestre][1]
			
	return aprovacoes_semestres
	
	

def posicao_turmaIngresso_semestral(df):
	iras = ira_semestral(df)
	iraMax = {}
	for matr in iras:
		for semestreAno in iras[matr]:
			if not (semestreAno in iraMax):
				iraMax[semestreAno] = iras[matr][semestreAno]
			else:
				if (iras[matr][semestreAno] > iraMax[semestreAno]):
					iraMax[semestreAno] = iras[matr][semestreAno]
	for matr in iras:
		for semestreAno in iras[matr]:
			iras[matr][semestreAno] /= iraMax[semestreAno]

	return iras


def periodo_real(df):
	aux = df.groupby(["MATR_ALUNO"])
	students = {}
	for x in aux:
		students[x[0]] = None
	return students


def periodo_pretendido(df):
	aux = df.groupby(["MATR_ALUNO", "ANO_INGRESSO", "SEMESTRE_INGRESSO"])
	students = {}
	for x in aux:
		students[x[0][0]] = (CURRENT_YEAR - int(x[0][1])) * 2 + CURRENT_SEMESTER - int(x[0][2]) + 1
	return students


def ira_semestral(df):
	aux = ira_por_quantidade_disciplinas(df)
	for matr in aux:
		for periodo in aux[matr]:
			aux[matr][periodo] = aux[matr][periodo][0]
	return aux


def ira_por_quantidade_disciplinas(df):
	students = {}
	df = df.dropna(subset=["MEDIA_FINAL"])

	total_students = len(df["MATR_ALUNO"])
	for i in range(total_students):
		matr = df["MATR_ALUNO"][i]
		if (not (matr in students)):
			students[matr] = {}

		ano = str(int(df["ANO"][i]))
		semestre = str(df["PERIODO"][i])
		situacao = int(df["SITUACAO"][i])
		nota = float(df["MEDIA_FINAL"][i])
		carga = float(df["CH_TOTAL"][i])
		#media_credito = int(df["MEDIA_CREDITO"][i])
		
		
		#if (situacao in Situation.SITUATION_AFFECT_IRA and media_credito != 0):
		if (situacao in Situation.SITUATION_AFFECT_IRA):
			if not (ano + "/" + semestre in students[matr]):
				students[matr][ano + "/" + semestre] = [0, 0, 0]
			
			students[matr][ano + "/" + semestre][0] += nota*carga
			students[matr][ano + "/" + semestre][1] += 1
			students[matr][ano + "/" + semestre][2] += carga

	for matr in students:
		for periodo in students[matr]:
			if (students[matr][periodo][2] != 0):
				students[matr][periodo][0] /= students[matr][periodo][2] * 100
	
	return (students)


def indice_aprovacao_semestral(df):
	students = {}
	df = df.dropna(subset=['MEDIA_FINAL'])
	total_students = len(df["MATR_ALUNO"])
	for i in range(total_students):
		matr = (df["MATR_ALUNO"][i])
		if (not (matr in students)):
			students[matr] = {}

		ano = str(int(df["ANO"][i]))
		semestre = str(df["PERIODO"][i])
		situacao = int(df["SITUACAO"][i])

		if not (ano + "/" + semestre in students[matr]):
			students[matr][ano + "/" + semestre] = [0, 0]

		if situacao in Situation.SITUATION_PASS:
			students[matr][ano + "/" + semestre][0] += 1
			students[matr][ano + "/" + semestre][1] += 1
		if situacao in Situation.SITUATION_FAIL:
			students[matr][ano + "/" + semestre][1] += 1
		
	return (students)


def aluno_turmas(df):
	students = {}
	df = df.dropna(subset=['MEDIA_FINAL'])

	situations = dict(Situation.SITUATIONS)

	for matr, hist in df.groupby('MATR_ALUNO'):
		students[matr] = []

		for _, row in hist.iterrows():
			data = {
				'ano': str(int(row["ANO"])),
				'codigo': row["COD_ATIV_CURRIC"],
				'nome': row["NOME_ATIV_CURRIC"],
				'nota': row["MEDIA_FINAL"],
				'semestre': row["PERIODO"],
				'situacao': situations.get(row["SITUACAO"], Situation.SIT_OUTROS)
			}

			students[matr].append(data)

	return students