Skip to content

Instantly share code, notes, and snippets.

@thcipriani

thcipriani/cashflow.sh Secret

Created November 13, 2024 02:13
Download ZIP

Revisions

  1. thcipriani created this gist Nov 13, 2024.
    80 changes: 80 additions & 0 deletions cashflow.sh
    View file
    Original file line number Diff line number Diff line change
    @@ -0,0 +1,80 @@
    #!/usr/bin/env bash

    # cashflow.sh - Plot cashflow using gnuplot
    # Usage: cashflow.sh
    # Dependencies: hledger, gnuplot, awk
    # Author: <https://www.sundialdreams.com/report-scripts-for-ledger-cli-with-gnuplot/> + minor modifications

    set -euo pipefail

    if ! command -v gnuplot &> /dev/null; then
    echo "gnuplot is not installed"
    exit 1
    fi

    income=$(mktemp)
    expense=$(mktemp)
    last_year=$(date --date "last year" +%Y)
    this_year=$(date +%Y)

    trap 'rm -f "$income" "$expense"' EXIT

    hledger register income -b "${this_year}-01-01" -M --output-format csv | awk -F, '
    NR > 1 {
    gsub(/"/, "", $2);
    gsub(/[\$,"]/, "", $7);
    amount = ($7 < 0 ? -$7 : $7);
    totals[$2] = amount;
    }
    END {
    for (date in totals)
    printf "%s %.2f\n", date, totals[date];
    }
    ' | sort > "$income"

    hledger register expenses -b "${this_year}-01-01" -M --output-format csv | awk -F, '
    NR > 1 {
    gsub(/"/, "", $2);
    gsub(/[\$,"]/, "", $7);
    amount = ($7 < 0 ? -$7 : $7);
    totals[$2] = amount;
    }
    END {
    for (date in totals)
    printf "%s %.2f\n", date, totals[date];
    }
    ' | sort > "$expense"

    cat <(echo "INCOME") "$income" <(echo "EXPENSE") "$expense"

    (cat <<EOF) | gnuplot
    set terminal qt size 1280,720 persist
    big_font = "League Spartan,12"
    small_font = "Source Code Pro,9"
    text_color = "#444444"
    border_color = "#f2f2f2"
    text_formatting(x, y) = (strlen(x) > 0) ? sprintf('%s: $%.0f', x, y) : sprintf('$%.0f', y)
    boxwidth = 1.5
    stats '$expense' using 2 nooutput name "expense"
    stats '$income' using 2 nooutput name "income"
    overall_max = (int(expense_max) > int(income_max)) ? int(expense_max) : int(income_max)
    set xdata time
    set key left top nobox font small_font textcolor rgb text_color
    set timefmt "%Y-%m-%d"
    set xrange ["${last_year}-12-20":"${this_year}-12-10"]
    set yrange [0:overall_max * 1.05]
    set xtics nomirror "$(date +%Y)-01-01",2592000 format "%b"
    unset mxtics
    set border lw 1 lc rgb border_color
    set xtics nomirror font small_font textcolor rgb text_color
    set title "INCOME VS EXPENSES\n{/*0.75(cumulative)" font big_font textcolor rgb text_color
    set style fill transparent solid 0.4 noborder
    set noylabel
    unset ytic
    unset ylabel
    plot "$income" using 1:2 with filledcurves x1 title "Income" linecolor rgb "seagreen", \
    '' using 1:2:(text_formatting('', column(2))) with labels font small_font offset 0,0.5 textcolor linestyle 0 notitle, \
    "$expense" using 1:2 with filledcurves y1=0 title "Expenses" linecolor rgb "light-salmon", \
    '' using 1:2:(text_formatting('', column(2))) with labels font small_font offset 0,-0.5 textcolor linestyle 0 notitle
    EOF
    101 changes: 101 additions & 0 deletions expenses.R
    View file
    Original file line number Diff line number Diff line change
    @@ -0,0 +1,101 @@
    # Load required libraries
    library(ggplot2)
    library(stringr)
    library(readr)
    library(dplyr)
    library(tidyr)
    library(reshape2)
    library(scales)
    library(systemfonts)
    library(showtext)

    font_family <- "roboto"
    showtext_auto()
    font_add_google("Roboto", "roboto")
    font_add_google("Redacted Script", "redacted")

    # last full month
    last_month <- as.character(as.Date(cut(Sys.Date(), "month")) - 1)
    last_month_full_name <- format(as.Date(last_month), "%B %Y")

    # Drop the day from the iso date
    # 2024-10-01 -> 2024-10
    last_month_col <- substr(last_month, 1, 7)

    # Load and reshape the data
    data <- read_csv("expenses.csv")

    data <- data %>% melt(id.vars = c("account", "average"), variable.name = "month", value.name = "expense")

    # Strip '$'
    data$expense <- gsub("\\$", "", data$expense)
    data$average <- gsub("\\$", "", data$average)

    # Convert float
    data$expense <- as.numeric(data$expense)
    data$average <- as.numeric(data$average)

    data$account <- str_to_title(gsub("expenses:", "", data$account))

    # Drop the "Unknown" row
    data <- data %>% filter(account != "Unknown")
    data <- data %>% filter((month == last_month_col))
    data <- data %>% arrange(desc(expense))

    # Rename the month to the month short name
    data$month <- sprintf("%s-%s", data$month, "01")
    data$month <- format(as.Date(data$month), "%b %Y")

    head(data)

    data <- data[1:4,]
    data$account <- factor(data$account, levels = rev(data$account))

    scale <- max(data$expense) * 1.1
    ten_percent <- scale * 0.005
    col_left_pos <- scale - scale * 0.3
    col_right_pos <- scale - scale * 0.2

    options(repr.plot.height=4)

    width = 0.8
    ggplot(data, aes(x = account, y = expense)) +
    geom_col(fill = "steelblue", alpha=0.1, width = 1) +
    geom_text(aes(label = account, y = ten_percent*2), hjust = "inward", size = 8) +
    geom_text(aes(label = label_dollar(accuracy=1)(expense), y = col_right_pos), hjust = 0, size = 6.25, color = "#222222") +
    geom_text(aes(label = label_dollar(accuracy=1)(average), y = col_left_pos), hjust = 0, size = 6, color = "#333333") +
    # geom_text(aes(label = '$0000.00', y = col_right_pos), hjust = 0, size = 6.25, color = "#222222", family="redacted") +
    # geom_text(aes(label = '$0000.00', y = col_left_pos), hjust = 0, size = 6, color = "#333333", family="redacted") +

    geom_errorbar(aes(ymin=0, ymax=0), color = "steelblue", linewidth=1.1) +

    scale_linetype_manual(values = c("Average" = "dotted")) +
    geom_errorbar(aes(ymin=average, ymax=average, group=account, linetype="Average"), color = "steelblue", show.legend=TRUE) +

    annotate("text", label = 'CATEGORY', x=4.75, y = ten_percent*2, hjust = "inward", size = 3, fontface = "bold") +
    annotate("text", label = 'AMOUNT', x =4.75, y = col_right_pos, hjust = 0, size = 3, fontface = "bold") +
    annotate("text", label = 'AVG', x = 4.75, y = col_left_pos + 20, hjust=0, size = 3, fontface = "bold") +

    scale_y_discrete(expand = c(0, 0)) +
    coord_flip(clip="off") +
    labs(title = "Expenses",
    subtitle = last_month_full_name,
    linetype = "",
    x = NULL, y = NULL) +
    theme_minimal() +
    theme(
    legend.position = c(0, -0.15),
    legend.justification = c(0, 0),
    legend.title = element_blank(),
    legend.text = element_text(size = 10, family = font_family),
    text = element_text(family = font_family, color="#212121"),
    plot.title = element_text(vjust = 2, size = 32, face = "bold"),
    plot.subtitle = element_text(vjust = 3.75, size = 18),
    axis.text.x = element_blank(),
    axis.ticks.x = element_blank(),
    panel.grid = element_blank(),
    axis.text.y = element_blank(),
    aspect.ratio = 0.2,
    plot.margin = margin(t = 1, r = 15, b = 20, l = 15),
    )
    ggsave("expenses_histogram.svg", width = 15, height = 4.75, limitsize = FALSE)
    7 changes: 7 additions & 0 deletions expenses.sh
    View file
    Original file line number Diff line number Diff line change
    @@ -0,0 +1,7 @@
    #!/usr/bin/env bash
    # Get the hledger data
    hledger bal -b 2024-01-01 -e 2024-11-01 expenses -M --depth=2 --no-total --average -O csv > expenses.csv

    Rscript expenses.R

    echo "Plot saved as expenses_histogram.svg"
    209 changes: 209 additions & 0 deletions plot-subaccounts.sh
    View file
    Original file line number Diff line number Diff line change
    @@ -0,0 +1,209 @@
    #!/usr/bin/env bash

    # plot-subaccounts.sh
    # --------------------
    # Setup: python3 -m virtualenv .venv && . .venv && pip install pandas
    # Author: Tyler Cipriani, 2024
    # License: GPLv3

    set -eo pipefail

    BASEDIR=$(cd "$(dirname "$0")" && pwd)
    VENVDIR="${BASEDIR}/../.venv/bin"

    show_help() {
    cat<<USE
    USAGE:
    $0 [OPTIONS] [OUTPUT_FILE]
    OPTIONS:
    -m, --months Number of months to include in the chart (default: 5)
    -a, --account Account to chart (default: expenses)
    -d, --depth Depth of the account to chart (default: 2)
    EXAMPLE:
    $0 -m 3 expenses.png
    USE
    }

    die() {
    printf '[ERROR] %s\n' "$1" >&2
    exit 1
    }


    hledger_bal() {
    local last_month months_ago account depth
    last_month=$1
    months_ago=$2
    account=$3
    depth=$4

    # Get the expenses for the last $months_ago months
    hledger bal "$account" -b "$months_ago" -e "$last_month" -M --depth="$depth" -O csv --no-total --average | "${VENVDIR}"/python3 -c "
    import sys
    import pandas as pd
    df = pd.read_csv(sys.stdin)
    df['account'] = df['account'].str.split(':').str[-1].str.title()
    for col in df.columns:
    if col != 'account':
    df[col] = df[col].str.replace(',', '').str.replace('\$', '').astype(float)
    df = df.sort_values(by='average', ascending=False)
    # df.drop(columns=['average'], inplace=True)
    df = df.drop(columns=['average'])
    # Pivot the data
    df_pivoted = df.set_index('account').T
    df_pivoted = df_pivoted.reset_index().rename(columns={'index': 'date'})
    sorted_columns = ['date'] + list(df['account'])
    df_pivoted = df_pivoted[sorted_columns]
    df_pivoted.to_csv(sys.stdout, index=False)
    "
    }

    plot_balances() {
    local data_file plot_file term num_accounts layout_cols layout_rows layout
    data_file="$1"
    shift
    if [ -z "$1" ]; then
    term='qt size 1280,720 persist'
    plot_file=
    else
    term='pngcairo size 1280,720'
    plot_file="set output '$1'"
    fi

    num_accounts=$(head -n 1 "$data_file" | awk -F, '{print NF - 1}')

    layout_cols=$(echo "sqrt($num_accounts)" | bc)
    layout_rows=$layout_cols

    while ((layout_cols * layout_rows < num_accounts)); do
    layout_cols=$((layout_cols + 1))
    if ((layout_cols * layout_rows < num_accounts)); then
    layout_rows=$((layout_rows + 1))
    fi
    done

    layout="$layout_cols,$layout_rows"

    # Gnuplot script
    (cat <<EOF) | gnuplot
    set datafile separator ","
    set key autotitle columnhead
    unset key
    set terminal $term
    $plot_file
    big_font = "League Spartan,12"
    small_font = "Source Code Pro,9"
    text_color = "#444444"
    border_color = "#f2f2f2"
    text_formatting(x, y) = (strlen(x) > 0) ? sprintf('%s: $%.0f', x, y) : sprintf('$%.0f', y)
    boxwidth = 1.5
    array mean[$num_accounts]
    array max[$num_accounts]
    do for [i=2:$num_accounts+1] {
    stats '$data_file' using i nooutput
    mean[i-1] = STATS_mean
    max[i-1] = STATS_max
    }
    overall_max = 0
    do for [i=1:$num_accounts] {
    overall_max = max[i] > overall_max ? max[i] : overall_max
    }
    set style fill transparent solid 0.4 noborder
    set style data histogram
    set style histogram clustered gap 1
    set boxwidth 1.5
    set multiplot layout $layout
    ymax = overall_max * 1.33
    print overall_max
    print ymax
    set yrange [0:ymax]
    set noylabel
    set xtics nomirror
    set ytics nomirror
    set border lw 1 lc rgb border_color
    set xtics font small_font textcolor rgb text_color
    unset ylabel
    unset ytic
    set xrange [-2:$months+1]
    do for [i=2:$num_accounts+1] {
    column_title = word(system(sprintf("head -n 1 %s | awk -F, '{print toupper(\$%d)}'", "$data_file", i)), 1)
    set label 1 column_title font big_font at graph 0.03, graph 0.9 left tc rgb text_color
    set label 2 text_formatting('Avg', mean[i-1]) font small_font at graph 0.03, mean[i-1]/ymax offset 0,0.5 tc rgb text_color
    plot '$data_file' using i:xtic(1) with histogram title col ls i-1, \
    '' using 0:i:(text_formatting('', column(i))) w labels font small_font offset 0,0.5 noti, \
    mean[i-1] noti with lines lw 1 dt 3
    }
    unset multiplot
    EOF
    }
    main () {
    if ! command -v gnuplot > /dev/null 2>&1; then
    die "gnuplot is not installed"
    fi
    if ! command -v csvsql > /dev/null 2>&1; then
    die "csvkit is not installed"
    fi
    if ! command -v "${VENVDIR}"/python3 > /dev/null 2>&1; then
    die "python3 virtualenv is not initialized"
    fi
    local months depth account output_file data_file
    months=6
    depth=2
    account=expenses
    output_file=
    while [ -n "$1" ]; do
    case "$1" in
    --months|-m)
    shift
    months="$1"
    ;;
    --account|-a)
    shift
    account="$1"
    ;;
    --depth|-d)
    shift
    depth="$1"
    ;;
    --help|-h)
    show_help
    exit 0
    ;;
    *)
    output_file="$1"
    ;;
    esac
    shift
    done
    months=$(( months - 1 ))
    local last_month months_ago
    last_month=$(date -d "$(date +'%Y-%m-01') -1 day" +'%Y-%m-%d')
    months_ago=$(date -d "${last_month} -$months months" +'%Y-%m-01')
    data_file=$(mktemp)
    hledger_bal "$last_month" "$months_ago" "$account" "$depth" > "$data_file"
    cat "$data_file"
    plot_balances "$data_file" "$output_file"
    if [ -n "$output_file" ]; then
    echo "Chart saved to $output_file"
    fi
    rm -f "$data_file"
    }
    main "$@"