Tree map

Last updated 2020-02-12 Read time 3 min

A treemap is a diagram that can be used to visualize numerical data with hierarchical categories. A typical example is a heat map of the Nikkei 225 or the S&P 500. This notebook uses squarify.

A treemap can also be created by using plotly.（ Treemap charts with Python - Plotly）

 1
 2
 3
 4
 5
 6
 7
 8
 9
10
11
12
13
14
15
import numpy as np
import matplotlib.pyplot as plt
import japanize_matplotlib
import squarify

np.random.seed(0)  # Fix random numbers

labels = ["A" * i for i in range(1, 5)]
sizes = [i * 10 for i in range(1, 5)]
colors = ["#%02x%02x%02x" % (i * 50, 0, 0) for i in range(1, 5)]

plt.figure(figsize=(5, 5))
squarify.plot(sizes, color=colors, label=labels)
plt.axis("off")
plt.show()

Tree map figure

Visualize my portfolio #

Suppose I have data on the acquisition price and current price for each stock I own. From there, I would create a heatmap like finviz.

Suppose we read the following data from csv.

※The data shown here are fictitious.

 1
 2
 3
 4
 5
 6
 7
 8
 9
10
11
12
13
14
15
16
17
import pandas as pd

data = [
    ["PBR", 80.20, 130.00],
    ["GOOG", 1188.0, 1588.0],
    ["FLNG", 70.90, 230.00],
    ["ZIM", 400.22, 630.10],
    ["GOGL", 120.20, 90.90],
    ["3466\nラサールロジ", 156.20, 147.00],  # 日本語表示のテスト用
]

df = pd.DataFrame(data)
df.columns = ["銘柄名", "取得価額", "現在の価額"]
df["評価損益"] = df["現在の価額"] - df["取得価額"]
df.head(6)

plt.show()

	銘柄名	取得価額	現在の価額	評価損益
0	PBR	80.20	130.0	49.80
1	GOOG	1188.00	1588.0	400.00
2	FLNG	70.90	230.0	159.10
3	ZIM	400.22	630.1	229.88
4	GOGL	120.20	90.9	-29.30
5	3466\nラサールロジ	156.20	147.0	-9.20

Specify the color of the treemap. #

Green for profitable areas and red for loss areas.

 1
 2
 3
 4
 5
 6
 7
 8
 9
10
11
12
13
14
15
16
17
18
19
20
21
colors = []
percents = []
for p_or_l, oac in zip(df["評価損益"], df["取得価額"]):
    percent = p_or_l / oac * 100

    if p_or_l > 0:
        g = np.min([percent * 255 / 100 + 100, 255.0])
        color = "#%02x%02x%02x" % (0, int(g), 0)
        colors.append(color)
    else:
        r = np.min([-percent * 255 / 100 + 100, 255])
        color = "#%02x%02x%02x" % (int(r), 0, 0)
        colors.append(color)

    percents.append(percent)

print(df["銘柄名"].values)
print(colors)
print(percents)

plt.show()

['PBR' 'GOOG' 'FLNG' 'ZIM' 'GOGL' '3466\nラサールロジ']
['#00ff00', '#00b900', '#00ff00', '#00f600', '#a20000', '#730000']
[62.094763092269325, 33.670033670033675, 224.4005641748942, 57.43840887511868, -24.376039933444257, -5.8898847631241935]

Displaying a treemap #

Let’s display the profit/loss in colors and the percentage of profit/loss on the treemap. Japanese characters are not garbled because import japanize_matplotlib is used at the beginning.

 1
 2
 3
 4
 5
 6
 7
 8
 9
10
11
current_prices = [cp for cp in df["現在の価額"]]
labels = [
    f"{name}\n{np.round(percent, 2)}％".replace("-", "▼")
    for name, percent in zip(df["銘柄名"], percents)
]

plt.figure(figsize=(10, 10))
plt.rcParams["font.size"] = 18
squarify.plot(current_prices, color=colors, label=labels)
plt.axis("off")
plt.show()

Japanese characters are not garbled because import japanize_… figure

Add cache display to the treemap #

Let’s also add a cache display to the treemap. The color should be gray.

1
2
3
4
5
6
7
plt.figure(figsize=(10, 10))
plt.rcParams["font.size"] = 18
squarify.plot(
    current_prices + [3500], color=colors + ["#ccc"], label=labels + ["キャッシュ"]
)
plt.axis("off")
plt.show()

The color should be gray figure

Donut Chart — Visually represent proportions with a circle