4. Intro to matplotlib¶

4.1. Intro¶

4.1.1. fig, ax, & plt.subplots()¶

import matplotlib.pyplot as plt
import pandas as pd

先來介紹 oo 的寫法
一率先用 fig, ax = plt.subplots() 當開頭
- fig 是 container which hold everything you see on the page
- ax 是 part of the page that holds data, it is the canvas
所以，在我們還沒有 assign data 給 ax 前，畫出來的圖，就是白的：

fig, ax = plt.subplots() # subplots 裡面沒寫參數，預設就是 1 ，表示你只要畫 1 張 subplot
plt.show()

../_images/datacamp_matplotlib_intro_5_0.png

接著，我們把 data 加進 ax 裡面

seattle_weather = pd.read_csv("data/seattle_weather.csv")
sub_data = seattle_weather.loc[seattle_weather.STATION == "USC00456295"]
sub_data.loc[:, ["DATE","MLY-PRCP-NORMAL"]] # prcp = precipitation (inches) 的縮寫，雨量

	DATE	MLY-PRCP-NORMAL
0	1	11.03
1	2	7.74
2	3	9.08
3	4	7.37
4	5	6.39
5	6	5.34
6	7	2.55
7	8	2.56
8	9	3.95
9	10	7.29
10	11	12.58
11	12	9.85

adding data to axes

fig, ax = plt.subplots()
ax.plot(sub_data["DATE"], sub_data["MLY-PRCP-NORMAL"])

[<matplotlib.lines.Line2D at 0x120efd5b0>]

../_images/datacamp_matplotlib_intro_9_1.png

那如果我想在這張圖上，再加上其他的 line ，我該怎麼做呢？
答案是：一直 ax.plot() 下去就好
例如剛剛這條線，我在資料中，還有 PR25, PR75 的資料，所以我可以畫出這個區間:

fig, ax = plt.subplots()
ax.plot(sub_data["DATE"], sub_data["MLY-PRCP-NORMAL"])
ax.plot(sub_data["DATE"], sub_data["MLY-PRCP-25PCTL"])
ax.plot(sub_data["DATE"], sub_data["MLY-PRCP-75PCTL"])

[<matplotlib.lines.Line2D at 0x1210de190>]

../_images/datacamp_matplotlib_intro_11_1.png

不錯，有上下限了，但不太美觀，下一節開始 customize 一些細節

4.1.2. customize¶

但因為這張圖，資料點其實只出現在 “月份” 上，所以最好加上 marker ，讓讀者知道資料出現在哪

fig, ax = plt.subplots()
ax.plot(sub_data["DATE"], sub_data["MLY-PRCP-NORMAL"], marker = "o")
ax.plot(sub_data["DATE"], sub_data["MLY-PRCP-25PCTL"], marker = "o")
ax.plot(sub_data["DATE"], sub_data["MLY-PRCP-75PCTL"], marker = "o")

[<matplotlib.lines.Line2D at 0x1210f08b0>]

../_images/datacamp_matplotlib_intro_15_1.png

我們也可以改變線條的 linestyle:

fig, ax = plt.subplots()
ax.plot(sub_data["DATE"], sub_data["MLY-PRCP-NORMAL"], marker = "o")
ax.plot(sub_data["DATE"], sub_data["MLY-PRCP-25PCTL"], marker = "o", linestyle = "--")
ax.plot(sub_data["DATE"], sub_data["MLY-PRCP-75PCTL"], marker = "o", linestyle = "--")

[<matplotlib.lines.Line2D at 0x1211a8850>]

../_images/datacamp_matplotlib_intro_17_1.png

也可以改顏色

fig, ax = plt.subplots()
ax.plot(sub_data["DATE"], sub_data["MLY-PRCP-NORMAL"], marker = "o", color = "blue")
ax.plot(sub_data["DATE"], sub_data["MLY-PRCP-25PCTL"], marker = "o", color = "blue", linestyle = "--")
ax.plot(sub_data["DATE"], sub_data["MLY-PRCP-75PCTL"], marker = "o", color = "blue", linestyle = "--")

[<matplotlib.lines.Line2D at 0x121214a30>]

../_images/datacamp_matplotlib_intro_19_1.png

改 x label, y label, title

fig, ax = plt.subplots()
ax.plot(sub_data["DATE"], sub_data["MLY-PRCP-NORMAL"], marker = "o", color = "blue")
ax.plot(sub_data["DATE"], sub_data["MLY-PRCP-25PCTL"], marker = "o", color = "blue", linestyle = "--")
ax.plot(sub_data["DATE"], sub_data["MLY-PRCP-75PCTL"], marker = "o", color = "blue", linestyle = "--")
ax.set_xlabel("Time (months)");
ax.set_ylabel("Precipitation (inches)")
ax.set_title("Weather in Seattle");

../_images/datacamp_matplotlib_intro_21_0.png

4.1.3. small multiples¶

fig, ax = plt.subplots(3, 2)
plt.show()

../_images/datacamp_matplotlib_intro_23_0.png

此時，ax 變成 array object，可以確認看看：

ax.shape

(3, 2)

所以，我等等要畫圖時，我要指名我要對 ax 這個 array 中的哪個 ax 畫圖
例如，我要對 (0,0) 畫圖，那就要寫 ax[0,0].plot()

fig, ax = plt.subplots(3,2)
ax[0,0].plot(sub_data["DATE"], sub_data["MLY-PRCP-NORMAL"], marker = "o", color = "blue")
ax[0,0].plot(sub_data["DATE"], sub_data["MLY-PRCP-25PCTL"], marker = "o", color = "blue", linestyle = "--")
ax[0,0].plot(sub_data["DATE"], sub_data["MLY-PRCP-75PCTL"], marker = "o", color = "blue", linestyle = "--")
ax[0,0].set_xlabel("Time (months)");
ax[0,0].set_ylabel("Precipitation (inches)")
ax[0,0].set_title("Weather in Seattle");

../_images/datacamp_matplotlib_intro_27_0.png

要注意的是，如果你的 subplot 只有一行或一列，那 ax 會退化到 1d array (而不是剛剛的 2d array)

fig, ax = plt.subplots(2,1) # 畫出 2 列 1 行
ax.shape

fig, ax = plt.subplots(1,2) # 畫出 1 列 2 行
ax.shape

(2,)

../_images/datacamp_matplotlib_intro_30_1.png

所以在畫圖時，就要用 ax[0].plot, ax[1].plot 來指定畫圖
舉例來說，我取兩個城市的資料來畫圖：

seattle_weather = pd.read_csv("data/seattle_weather.csv")
austin_weather = pd.read_csv("data/austin_weather.csv")

seattle_sub = seattle_weather.loc[seattle_weather.STATION == "USC00456295"]
austin_sub = austin_weather.loc[austin_weather.STATION == "USW00013904"]

fig, ax = plt.subplots(2,1) # 畫出 2 列 1 行

ax[0].plot(seattle_sub["DATE"], seattle_sub["MLY-PRCP-NORMAL"], marker = "o", color = "blue")
ax[0].plot(seattle_sub["DATE"], seattle_sub["MLY-PRCP-25PCTL"], marker = "o", color = "blue", linestyle = "--")
ax[0].plot(seattle_sub["DATE"], seattle_sub["MLY-PRCP-75PCTL"], marker = "o", color = "blue", linestyle = "--")

ax[1].plot(austin_sub["DATE"], austin_sub["MLY-PRCP-NORMAL"], marker = "o", color = "red")
ax[1].plot(austin_sub["DATE"], austin_sub["MLY-PRCP-25PCTL"], marker = "o", color = "red", linestyle = "--")
ax[1].plot(austin_sub["DATE"], austin_sub["MLY-PRCP-75PCTL"], marker = "o", color = "red", linestyle = "--")

ax[0].set_ylabel("Precipitation (inches)");
ax[1].set_ylabel("Precipitation (inches)");

../_images/datacamp_matplotlib_intro_33_0.png

雖然累了點，很多重複性高的 code，但結果跟預期差不多
但我們有發現， y 軸的 scale 不同，所以，我可以用 plt.subplots(2, 1, sharey = True)，來強制 y 軸做 share (相同 scale)

fig, ax = plt.subplots(2,1, sharey = True) # 畫出 2 列 1 行

ax[0].plot(seattle_sub["DATE"], seattle_sub["MLY-PRCP-NORMAL"], marker = "o", color = "blue")
ax[0].plot(seattle_sub["DATE"], seattle_sub["MLY-PRCP-25PCTL"], marker = "o", color = "blue", linestyle = "--")
ax[0].plot(seattle_sub["DATE"], seattle_sub["MLY-PRCP-75PCTL"], marker = "o", color = "blue", linestyle = "--")

ax[1].plot(austin_sub["DATE"], austin_sub["MLY-PRCP-NORMAL"], marker = "o", color = "red")
ax[1].plot(austin_sub["DATE"], austin_sub["MLY-PRCP-25PCTL"], marker = "o", color = "red", linestyle = "--")
ax[1].plot(austin_sub["DATE"], austin_sub["MLY-PRCP-75PCTL"], marker = "o", color = "red", linestyle = "--")

ax[0].set_ylabel("Precipitation (inches)");
ax[1].set_ylabel("Precipitation (inches)");

../_images/datacamp_matplotlib_intro_35_0.png

4.2. Time- series data¶

climate_change = pd.read_csv(
    "data/climate_change.csv", 
    parse_dates=["date"], 
    index_col = "date"
)
climate_change

	co2	relative_temp
date
1958-03-06	315.71	0.10
1958-04-06	317.45	0.01
1958-05-06	317.50	0.08
1958-06-06	NaN	-0.05
1958-07-06	315.86	0.06
...	...	...
2016-08-06	402.27	0.98
2016-09-06	401.05	0.87
2016-10-06	401.59	0.89
2016-11-06	403.55	0.93
2016-12-06	404.45	0.81

706 rows × 2 columns

fig, ax = plt.subplots()

ax.plot(climate_change.index, climate_change.relative_temp)
ax.set_xlabel("Time")
ax.set_ylabel("Relative temperature (Celsius)")

Text(0, 0.5, 'Relative temperature (Celsius)')

../_images/datacamp_matplotlib_intro_38_1.png

把時間軸 zoom in 一下 (因為時間現在是 index，所以可以很方便的做 subset)

fig, ax = plt.subplots()
seventies = climate_change["1970-01-01":"1979-12-31"]
ax.plot(seventies.index, seventies["co2"])

[<matplotlib.lines.Line2D at 0x122da67c0>]

../_images/datacamp_matplotlib_intro_40_1.png

4.2.1. 雙軸圖¶

fig, ax = plt.subplots()
ax.plot(climate_change.index, climate_change["co2"], color = "blue")
ax.set_xlabel("Time")
ax.set_ylabel("CO2 (ppm)", color = "blue")
ax.tick_params("y", colors = "blue")

ax2 = ax.twinx()
ax2.plot(climate_change.index, climate_change["relative_temp"], color = "red")
ax2.set_ylabel("Relative temperature (Celsius)", color = "red")
ax2.tick_params("y", colors = "red")

../_images/datacamp_matplotlib_intro_42_0.png

來寫個 function，讓我們可以做得更快一點

def plot_timeseries(axes, x, y, color, xlabel, ylabel):
    axes.plot(x, y, color = color)
    axes.set_xlabel(xlabel)
    axes.set_ylabel(ylabel, color = color)
    axes.tick_params("y", colors = color)

fig, ax = plt.subplots()
plot_timeseries(
    axes = ax, 
    x = climate_change.index, 
    y = climate_change["co2"], 
    color = "blue",
    xlabel = "Time",
    ylabel = "CO2 (ppm)"
)

ax2 = ax.twinx()
plot_timeseries(
    axes = ax2, 
    x = climate_change.index, 
    y = climate_change["relative_temp"], 
    color = "red",
    xlabel = "Time",
    ylabel = "Relative temperature (Celsius)"
)

../_images/datacamp_matplotlib_intro_45_0.png

4.2.2. 加上 annotation¶

fig, ax = plt.subplots()
plot_timeseries(
    axes = ax, 
    x = climate_change.index, 
    y = climate_change["co2"], 
    color = "blue",
    xlabel = "Time",
    ylabel = "CO2 (ppm)"
)

ax2 = ax.twinx()
plot_timeseries(
    axes = ax2, 
    x = climate_change.index, 
    y = climate_change["relative_temp"], 
    color = "red",
    xlabel = "Time",
    ylabel = "Relative temperature (Celsius)"
)
ax2.annotate(">1 degree", # annotation 的文字
             xy=(pd.Timestamp("2015-10-06"), 1), # annotation 的點所在的座標
             xytext = (pd.Timestamp('2008-10-06'), -0.2), # 要顯示的文字所在的座標
             arrowprops={"arrowstyle": "->", # 箭頭，從文字指向要標記的點
                         "color": "gray"})

Text(2008-10-06 00:00:00, -0.2, '>1 degree')

../_images/datacamp_matplotlib_intro_47_1.png

4.3. Statistical plots¶

medals = pd.read_csv("data/medals_by_country_2016.csv", index_col = 0)
medals

	Bronze	Gold	Silver
United States	67	137	52
Germany	67	47	43
Great Britain	26	64	55
Russia	35	50	28
China	35	44	30
France	21	20	55
Australia	25	23	34
Italy	24	8	38
Canada	61	4	4
Japan	34	17	13

4.3.1. Bar chart¶

4.3.1.1. Simple bar chart¶

我們想畫出，各國的金牌數量

fig, ax = plt.subplots()
ax.bar(medals.index, medals["Gold"])

<BarContainer object of 10 artists>

../_images/datacamp_matplotlib_intro_53_1.png

xlabel 都重疊再一起，我們把它轉 90 度

fig, ax = plt.subplots();
ax.bar(medals.index, medals["Gold"]);
ax.set_xticklabels(medals.index, rotation = 90);
ax.set_ylabel("Number of medals");

/var/folders/j9/71c8r2vs343cb9329xbww0240000gn/T/ipykernel_71988/528635632.py:3: UserWarning: FixedFormatter should only be used together with FixedLocator
  ax.set_xticklabels(medals.index, rotation = 90);

../_images/datacamp_matplotlib_intro_55_1.png

4.3.1.2. stack bar chart¶

我們可以做 stack bar chart

fig, ax = plt.subplots();
ax.bar(medals.index, medals["Gold"], label = "Gold");
ax.bar(medals.index, medals["Silver"], bottom = medals["Gold"], label = "Silver")
ax.bar(medals.index, medals["Bronze"],
       bottom = medals["Gold"]+medals["Silver"], 
       label = "Bronze")

ax.set_xticklabels(medals.index, rotation = 90);
ax.set_ylabel("Number of medals");
ax.legend()

/var/folders/j9/71c8r2vs343cb9329xbww0240000gn/T/ipykernel_71988/3216558047.py:8: UserWarning: FixedFormatter should only be used together with FixedLocator
  ax.set_xticklabels(medals.index, rotation = 90);

<matplotlib.legend.Legend at 0x122fc2b80>

../_images/datacamp_matplotlib_intro_58_2.png

4.3.2. Histogram¶

tt = pd.read_csv("data/summer2016.csv", index_col = 0)
mens_rowing = tt.query("(Sex == 'M') & (Sport == 'Rowing')")
mens_gymnastic = tt.query("(Sex == 'M') & (Sport == 'Gymnastics')")

fig, ax = plt.subplots()
ax.hist(mens_rowing["Height"], label = "Rowing", bins = 5)
ax.hist(mens_gymnastic["Height"], label = "Gymnastics", bins = 5)
ax.set_xlabel("Height (cm)")
ax.set_ylabel("# of observations")
ax.legend()

<matplotlib.legend.Legend at 0x123066910>

../_images/datacamp_matplotlib_intro_61_1.png

4.3.3. Error bars¶

幫 bar chart 加上 error bar

fig, ax = plt.subplots()

ax.bar("Rowing", 
       mens_rowing["Height"].mean(),
       yerr = mens_rowing["Height"].std())
ax.bar(
    "Gymnastics",
    mens_gymnastic["Height"].mean(),
    yerr = mens_gymnastic["Height"].std()
)

<BarContainer object of 1 artists>

../_images/datacamp_matplotlib_intro_64_1.png

我們也可以用迴圈，畫完所有運動員的 bar chart with error bar

tt = pd.read_csv("data/summer2016.csv", index_col = 0)
sports = tt.Sport.unique()

fig, ax = plt.subplots()
fig.set_size_inches([10, 5])
for sport in sports:
    sport_df = tt.loc[tt.Sport == sport]
    ax.bar(sport, sport_df["Height"].mean(), yerr = sport_df["Height"].std())

ax.set_ylabel("Height (cm)")
ax.set_xticklabels(sports, rotation = 90)
plt.show()

/var/folders/j9/71c8r2vs343cb9329xbww0240000gn/T/ipykernel_71988/1536233426.py:11: UserWarning: FixedFormatter should only be used together with FixedLocator
  ax.set_xticklabels(sports, rotation = 90)

../_images/datacamp_matplotlib_intro_66_1.png

幫 line chart 加上 error bar

fig, ax = plt.subplots()

ax.errorbar(
    seattle_sub["DATE"],
    seattle_sub["MLY-TAVG-NORMAL"],
    yerr = seattle_sub["MLY-TAVG-STDDEV"],
    label = "Seattle"
)

ax.errorbar(
    austin_sub["DATE"],
    austin_sub["MLY-TAVG-NORMAL"],
    yerr = austin_sub["MLY-TAVG-STDDEV"],
    label = "Austin"
)

ax.set_ylabel("Temperature (Fahrenheit")
ax.set_xlabel("Month")
ax.legend()

<matplotlib.legend.Legend at 0x1234a0820>

../_images/datacamp_matplotlib_intro_68_1.png

4.3.4. Boxplot¶

fig, ax = plt.subplots()
ax.boxplot([mens_rowing["Height"], mens_gymnastic["Height"]])
ax.set_xticklabels(["Rowing", "Gymnastics"])
ax.set_ylabel("Height (cm)")

Text(0, 0.5, 'Height (cm)')

../_images/datacamp_matplotlib_intro_70_1.png

4.3.5. scatter plot¶

fig, ax = plt.subplots()
ax.scatter(climate_change["co2"], climate_change["relative_temp"])
ax.set_xlabel("CO2 (ppm)")
ax.set_ylabel("Relative temperature (Celsius)")

Text(0, 0.5, 'Relative temperature (Celsius)')

../_images/datacamp_matplotlib_intro_72_1.png

eighties = climate_change["1980-01-01":"1989-12-31"]
nineties = climate_change["1990-01-01":"1999-12-31"]

fig, ax = plt.subplots()
ax.scatter(eighties["co2"], eighties["relative_temp"],
           color = "red", label = "eighties")
ax.scatter(nineties["co2"], nineties["relative_temp"],
           color = "blue", label = "nineties")
ax.legend()

ax.set_xlabel("CO2 (ppm)")
ax.set_ylabel("Relative temperature (Celsius)")

Text(0, 0.5, 'Relative temperature (Celsius)')

../_images/datacamp_matplotlib_intro_73_1.png

fig, ax = plt.subplots()
ax.scatter(climate_change["co2"], climate_change["relative_temp"],
           c = climate_change.index)
ax.set_xlabel("CO2 (ppm)")
ax.set_ylabel("Relative temperature (Celsius)")

Text(0, 0.5, 'Relative temperature (Celsius)')

../_images/datacamp_matplotlib_intro_74_1.png

4.4. Saving files¶

fig, ax = plt.subplots()

# fig.savefig("gold_medals.png")
# fig.savefig("gold_medals.jpg")
# fig.savefig("gold_medals.svg")

fig, ax = plt.subplots()
fig.set_size_inches([5, 3]) # width, height

My sample book

Intro to matplotlib

Contents

4. Intro to matplotlib¶

4.1. Intro¶

4.1.1. fig, ax, & plt.subplots()¶

4.1.2. customize¶

4.1.3. small multiples¶

4.2. Time- series data¶

4.2.1. 雙軸圖¶

4.2.2. 加上 annotation¶

4.3. Statistical plots¶

4.3.1. Bar chart¶

4.3.1.1. Simple bar chart¶

4.3.1.2. stack bar chart¶

4.3.2. Histogram¶

4.3.3. Error bars¶

4.3.4. Boxplot¶

4.3.5. scatter plot¶

4.4. Saving files¶