an.tsouchlos/banking-breakdown
1
0
Fork 0
Code Issues 2 Pull Requests 1 Packages Projects Releases Wiki Activity

Add expense-by-category figure

Browse Source
This commit is contained in:
Andreas Tsouchlos 2024-01-06 22:42:46 +01:00
parent 99a7920118
commit 1e4efa0c5b
7 changed files with 257 additions and 90 deletions
Show Stats Download Patch File Download Diff File Expand all files Collapse all files

11
banking_breakdown/__main__.py
View File

@ -1,6 +1,5 @@
from banking_breakdown import document_builder
from banking_breakdown import statement_parser
from banking_breakdown import ui
from banking_breakdown import ui, regex_categorizer, statement_parser, \
document_builder
import argparse
@ -10,6 +9,9 @@ def categorize_func(args):
df = pd.read_csv(args.i, delimiter=args.d)
if args.f is not None:
df = regex_categorizer.assign_categories(df, args.f)
import signal
signal.signal(signal.SIGINT, signal.SIG_DFL)
@ -17,7 +19,8 @@ def categorize_func(args):
def report_func(args):
print("Report")
report_data = statement_parser.parse_statement(args.i)
document_builder.build_document(report_data)
#

2
banking_breakdown/document_builder.py
View File

@ -14,6 +14,8 @@ def _serialize_report_data(report_data: types.ReportData):
report_data.net_income.to_csv('build/net_income.csv', index=False)
report_data.category_overview.to_csv('build/category_overview.csv',
index=False)
report_data.expenses_by_category.to_csv('build/expenses_by_category.csv',
index=False)
report_data.total_value.to_csv('build/total_value.csv', index=False)
report_data.detailed_balance.to_csv('build/detailed_balance.csv',
index=False)

54
banking_breakdown/regex_categorizer.py Normal file
View File

@ -0,0 +1,54 @@
import pandas as pd
import json
def _is_str_column(s: pd.Series):
"""Check if the type of a pandas DataFrame column is str.
Taken from https://stackoverflow.com/a/67001213/3433817.
"""
if isinstance(s.dtype, pd.StringDtype):
# The series was explicitly created as a string series (Pandas>=1.0.0)
return True
elif s.dtype == 'object':
# Object series, check each value
return all((v is None) or isinstance(v, str) for v in s)
else:
return False
def _read_regex_dict(regex_file: str):
with open(regex_file, 'r') as f:
return json.load(f)
def assign_categories(df: pd.DataFrame, regex_file: str) -> pd.DataFrame:
if 'category' not in df.columns:
df['category'] = [' '] * len(df.index)
regex_dict = _read_regex_dict(regex_file)
df = df.fillna('')
for column in df.columns:
if not _is_str_column(df[column]):
continue
for category in regex_dict:
for regex in regex_dict[category]:
matched = df[column].str.contains(regex, regex=True)
df.loc[matched, 'category'] = category
return df
def main():
df = pd.read_csv('../res/bank_statement_2023_categorized.csv')
df = assign_categories(df, regex_file='../res/regexes.json')
print(df['category'])
if __name__ == "__main__":
main()

182
banking_breakdown/statement_parser.py
View File

@ -6,94 +6,104 @@ import re
import numpy as np
# def _read_regex_dict(regex_file: str = "res/category_regexes.json"):
# with open(regex_file, 'r') as f:
# return json.load(f)
#
#
# def _tag_with_category(df: pd.DataFrame) -> pd.DataFrame:
# regex_dict = _read_regex_dict()
#
# return df
#
#
# def _compute_total_balance(df: pd.DataFrame) -> pd.DataFrame:
# stripped_df = pd.DataFrame(
# {'t': df["Valutadatum"], 'value': df["Saldo nach Buchung"]})
#
# stripped_df.index = stripped_df['t']
# gb = stripped_df.groupby(pd.Grouper(freq='M'))
#
# result = gb.tail(1)['value'].reset_index()
#
# return result
#
#
# def _compute_net_income(df: pd.DataFrame) -> pd.DataFrame:
# stripped_df = pd.DataFrame({'t': df["Valutadatum"], 'value': df["Betrag"]})
#
# stripped_df.index = stripped_df['t']
# gb = stripped_df.groupby(pd.Grouper(freq='M'))
#
# result = gb["value"].sum().reset_index()
# return result
#
#
# def _compute_category_overview(df: pd.DataFrame) -> pd.DataFrame:
# categories = ["Social life", "Other", "Food", "Hobbies",
# "Rent \\& Utilities", "Education", "Transportation"]
# values = np.array([10, 12, 53, 12, 90, 23, 32])
# values = values / values.sum() * 100
# values = np.round(values, decimals=1)
# values[-1] += 100 - np.sum(values)
#
# category_overview_df = pd.DataFrame(
# {"category": categories, "value": values})
#
# return category_overview_df
#
#
# def _compute_detailed_balance(df: pd.DataFrame) -> pd.DataFrame:
# return pd.DataFrame({'t': df["Valutadatum"],
# 'value': df["Saldo nach Buchung"]})
#
#
# def parse_statement(filename: str) -> types.ReportData:
# df = pd.read_csv(filename, delimiter=';', decimal=",")
# df["Valutadatum"] = pd.to_datetime(df["Valutadatum"], format='%d.%m.%Y')
#
# category_overview_df = _compute_category_overview(df)
# total_balance_df = _compute_total_balance(df)
# net_income_df = _compute_net_income(df)
# detailed_balance_df = _compute_detailed_balance(df)
#
# return types.ReportData(category_overview_df,
# net_income_df,
# total_balance_df,
# detailed_balance_df)
#
#
# def main():
# report_data = parse_statement("../res/bank_statement_2023.csv")
#
#
# if __name__ == "__main__":
# main()
def _escape_string(to_escape: str):
return to_escape.translate(str.maketrans({"&": r"\&"}))
def get_stripped_statement(filename: str) -> pd.DataFrame:
# df = pd.read_csv(filename, delimiter=';', decimal=",")
df = pd.read_csv(filename, delimiter=';')
df["Valutadatum"] = (pd.to_datetime(df["Valutadatum"], format='%d.%m.%Y')
.dt.strftime('%Y-%m-%d'))
def _compute_total_balance(df: pd.DataFrame) -> pd.DataFrame:
stripped_df = pd.DataFrame(
{'t': df["t"], 'value': df["balance"]})
result = pd.DataFrame({'t': df["Valutadatum"],
'other party': df["Name Zahlungsbeteiligter"],
'value': df["Betrag"],
'balance': df["Saldo nach Buchung"],
'category': [''] * len(df["Valutadatum"]),
'description': df["Buchungstext"],
'purpose': df["Verwendungszweck"]
})
stripped_df.index = stripped_df['t']
gb = stripped_df.groupby(pd.Grouper(freq='M'))
result = gb.tail(1)['value'].reset_index()
return result
def _compute_net_income(df: pd.DataFrame) -> pd.DataFrame:
stripped_df = pd.DataFrame({'t': df["t"], 'value': df["value"]})
stripped_df.index = stripped_df['t']
gb = stripped_df.groupby(pd.Grouper(freq='M'))
result = gb["value"].sum().reset_index()
return result
def _compute_category_overview(df: pd.DataFrame) -> pd.DataFrame:
df = df.loc[df['value'] < 0]
df = df.drop('t', axis=1)
df = df.groupby(['category']).sum().reset_index()
values = (df['value'] / df['value'].sum() * 100).to_numpy()
values[-1] += 100 - np.sum(values)
values = np.round(values, decimals=1)
categories = [_escape_string(category) for category in df['category']]
category_overview_df = pd.DataFrame(
{"category": categories, "value": values})
category_overview_df = category_overview_df.sort_values('value',
ascending=False)
return category_overview_df
def _compute_expenses_by_category(complete_df: pd.DataFrame) -> pd.DataFrame:
complete_df = complete_df.loc[complete_df['value'] < 0]
complete_df['value'] = -complete_df['value']
complete_df.index = complete_df['t']
complete_gb = complete_df.groupby(pd.Grouper(freq='M'))
categories = complete_df['category'].unique()
data_dict = {category: [] for category in categories}
for (month_date, month_df) in complete_gb:
month_df = month_df.drop('t', axis=1).reset_index().drop('t', axis=1)
category_df = month_df.groupby(['category']).sum().reset_index()
for _, row in category_df.iterrows():
data_dict[row['category']].append(row['value'])
non_listed = list(set(categories) - set(category_df['category']))
for category in non_listed:
data_dict[category].append(0)
result = pd.DataFrame(data_dict)
result = result.reindex(result.mean().sort_values(ascending=False).index,
axis=1)
result['t'] = complete_gb.tail(1).drop('t', axis=1).reset_index()['t']
return result
def _compute_detailed_balance(df: pd.DataFrame) -> pd.DataFrame:
return pd.DataFrame({'t': df["t"],
'value': df["balance"]})
def parse_statement(filename: str) -> types.ReportData:
df = pd.read_csv(filename)
df["t"] = pd.to_datetime(df["t"], format='%Y-%m-%d')
category_overview_df = _compute_category_overview(df)
total_balance_df = _compute_total_balance(df)
net_income_df = _compute_net_income(df)
detailed_balance_df = _compute_detailed_balance(df)
expenses_by_category_df = _compute_expenses_by_category(df)
return types.ReportData(category_overview_df,
expenses_by_category_df,
net_income_df,
total_balance_df,
detailed_balance_df, )
def main():
report_data = parse_statement("../res/bank_statement_2023_categorized.csv")
if __name__ == "__main__":
main()

5
banking_breakdown/test.json Normal file
View File

@ -0,0 +1,5 @@
{
"asdf": [
"Kinemic"
]
}

1
banking_breakdown/types.py
View File

@ -5,6 +5,7 @@ import pandas as pd
@dataclass
class ReportData:
category_overview: pd.DataFrame
expenses_by_category: pd.DataFrame
net_income: pd.DataFrame
total_value: pd.DataFrame
detailed_balance: pd.DataFrame

92
res/report.tex
View File

@ -212,5 +212,97 @@
\end{figure}
\begin{figure}[H]
\centering
% Read table
\pgfplotstableread[col sep=comma]{expenses_by_category.csv}\expbycattable
\pgfplotstablegetcolsof{\expbycattable}
\pgfmathtruncatemacro\NumCols{\pgfplotsretval-1}
\begin{subfigure}[c]{\textwidth}
\centering
\begin{tikzpicture}
\begin{axis}[
stack plots=y,
area style,
date coordinates in=x,
width=\textwidth,
height=0.375\textwidth,
xticklabel=\month.\shortyear{\year},
xtick=data,
enlargelimits=false,
xticklabel style={
rotate=60,
anchor=near xticklabel,
},
legend columns=5,
legend style={at={(0.5,-0.6)},anchor=south},
ylabel={Expenses in €},
ymin=0,
]
% For each
\pgfplotsinvokeforeach{0,...,\NumCols/2 -1}{
% Define color
\pgfmathparse{1000 / (\NumCols/2 -1) * #1}
\extractcolormapcolor{tempcol#1}{\pgfmathresult}
% Add plot
\addplot+[tempcol#1]
table[col sep=comma, x=t, y index=#1]
{\expbycattable} \closedcycle;
% Add legend entry (https://tex.stackexchange.com/a/405018)
\pgfplotstablegetcolumnnamebyindex{#1}\of{\expbycattable}\to\pgfplotsretval
\expandafter\addlegendentry\expandafter{\pgfplotsretval}
}
\end{axis}
\end{tikzpicture}
\end{subfigure}\\[1em]
\begin{subfigure}[c]{\textwidth}
\centering
\begin{tikzpicture}
\begin{axis}[
stack plots=y,
area style,
date coordinates in=x,
width=\textwidth,
height=0.375\textwidth,
xticklabel=\month.\shortyear{\year},
xtick=data,
enlargelimits=false,
xticklabel style={
rotate=60,
anchor=near xticklabel,
},
legend columns=5,
legend style={at={(0.5,-0.6)},anchor=south},
ylabel={Expenses in €},
ymin=0,
]
% For each
\pgfplotsinvokeforeach{\NumCols/2,...,\NumCols-1}{
% Define color
\pgfmathparse{1000 * (#1 - \NumCols/2) / (\NumCols-1 - \NumCols/2)}
\extractcolormapcolor{tempcol#1}{\pgfmathresult}
% Add plot
\addplot+[tempcol#1]
table[col sep=comma, x=t, y index=#1]
{\expbycattable} \closedcycle;
% Add legend entry (https://tex.stackexchange.com/a/405018)
\pgfplotstablegetcolumnnamebyindex{#1}\of{\expbycattable}\to\pgfplotsretval
\expandafter\addlegendentry\expandafter{\pgfplotsretval}
}
\end{axis}
\end{tikzpicture}
\end{subfigure}
\caption{Expenses by category}
\end{figure}
\end{document}