def compute_applications(data, base_date):
previous_month_data = rs.records_for_month(data,
dth.months_ago(base_date, 1))
current_month_data = rs.records_for_month(data, base_date)
difference = rs.total_amount_by(
APPLICATIONS_GROUPBY, current_month_data) - rs.total_amount_by(
APPLICATIONS_GROUPBY, previous_month_data)
return difference.amount
def food_expenses(expenses, base_date):
data = rs.total_amount_by('category',
rs.records_for_month(expenses, base_date))
if 'mercado' in data.index and 'restaurante' in data.index:
return pd.Series(data.loc['mercado'].amount +
data.loc['restaurante'].amount) * -1
else:
return pd.Series(0)
def distribution(expenses, incomes, categories={}):
expenses_by_category = rs.total_amount_by('category', expenses)
dist_by_spent = distribution_by_category(expenses, expenses)
dist_by_income = distribution_by_category(
expenses, incomes[incomes.category == 'renda']) * -1
data = {
'amount #': expenses_by_category.amount,
'% by expenses': dist_by_spent.amount,
'% by income': dist_by_income.amount,
'% by income goal': categories,
}
return pd.DataFrame(data,
columns=list(data.keys())).sort_values('amount #')
def preprocess_applications(funds):
funds = normalize_amount(funds, 'application')
funds = rs.total_amount_by(['date', 'title', 'type', 'account', 'goal'],
funds).reset_index()
funds = rs.describe_over_time(funds, compute_applications)\
.transpose()\
.reset_index()\
.melt(id_vars=APPLICATIONS_GROUPBY,
var_name='date',
value_name='amount')
# Fix date to be in the beginning of month
funds['date'] = pd.to_datetime(
funds.date.map(lambda d: dth.beginning_of_month(rs.month_to_date(d))))
funds['category'] = 'aplicação'
# Remove first month because there is no record for its previous month
funds = funds[funds.date > funds.date.min()]
return funds
def expenses_by_month(expenses):
return rs.total_amount_by(rs.groupby_month(expenses), expenses)
def incomes_by_month(incomes):
income_salary = incomes[incomes.category == 'renda'][['date', 'amount']]
return rs.total_amount_by(rs.groupby_month(income_salary), income_salary)
def total_invested_by(column, invest):
"""Total invested grouped by column"""
return rs.total_amount_by(column, ft.invested(invest))
def applications_for_month(incomes, base_date):
"""Total applications on current month given base_date"""
applications = rs.records_for_month(ft.applications(incomes), base_date)
return rs.total_amount_by('title', applications)
def total_discounts_by(column, incomes):
"""Total discounts grouped by column"""
return rs.total_amount_by(column, ft.discounts(incomes))
def total_income_by(column, incomes):
"""Total incomes/salaries grouped by column"""
return rs.total_amount_by(column, ft.income(incomes))
def total_applications_by(column, invest):
"""Total applications grouped by column"""
return rs.total_amount_by(column, ft.applications(invest))
def distribution_by_category(expenses, denominator):
denominator_sum = denominator.amount.sum()
expenses_by_category = rs.total_amount_by('category',
expenses).sort_values('amount')
return expenses_by_category / denominator_sum
def preprocess_invested(funds):
funds = normalize_amount(funds, 'gross amount')
funds = rs.total_amount_by(['date', 'title', 'type', 'account', 'goal'],
funds).reset_index()
funds['category'] = 'valor aplicado'
return funds