数据可视化之美 -- 以Matlab、Python为工具

3 P8 h+ j; a+ l3 @9 V% f

# r' G$ O5 p3 c* S A5 ~

3 d5 f9 R: h7 _/ |0 c9 l0 r

在我们科研、工作中，将数据完美展现出来尤为重要。

& L' R; [6 r* z$ l

数据可视化是以数据为视角，探索世界。我们真正想要的是 — 数据视觉，以数据为工具，以可视化为手段，目的是描述真实，探索世界。

下面介绍一些数据可视化的作品（包含部分代码），主要是地学领域，可迁移至其他学科。

4 [% [, n" A9 {+ F8 i

Example 1 ：散点图、密度图（Python）

W/ [$ ~6 w' Y' `

import numpy as np

K% O8 Y- p1 A0 k4 H& H

import matplotlib.pyplot as plt

5 s- w( G: H# e4 v. `' D4 X

# 创建随机数

6 o: W: j7 [1 D) I4 O0 ?7 F

n = 100000

x = np.random.randn(n)

- z' D7 U1 R, {: B

y = (1.5 * x) + np.random.randn(n)

9 Z$ Z, }0 g! G; E

fig1 = plt.figure()

! j& {) a5 F% h# r

plt.plot(x,y,.r)

plt.xlabel(x)

* a6 A8 u' d/ p: }' k3 e( g

plt.ylabel(y)

plt.savefig(2D_1V1.png,dpi=600)

* ]5 ~5 z1 `" ]9 g* ^

nbins = 200

! y' D! D+ o' |% O5 o; }2 B

H, xedges, yedges = np.histogram2d(x,y,bins=nbins)

5 Z* _' l$ J7 A% J$ \

# H needs to be rotated and flipped

H = np.rot90(H)

( X# r" O0 `( ~6 I, a; _

H = np.flipud(H)

# 将zeros mask

Hmasked = np.ma.masked_where(H==0,H)

) v/ J9 a! i( l6 e) Y0 ?/ @

# Plot 2D histogram using pcolor

* h# f' E4 t: c- q# v- ?

fig2 = plt.figure()

9 E$ G+ ]$ i0 k J

plt.pcolormesh(xedges,yedges,Hmasked)

plt.xlabel(x)

% A* d! y+ m) G+ P" @" ?

plt.ylabel(y)

cbar = plt.colorbar()

cbar.ax.set_ylabel(Counts)

plt.savefig(2D_2V1.png,dpi=600)

3 k6 L3 `: q2 s

plt.show()

打开凤凰新闻，查看更多高清图片

8 }+ U3 ]; r* ^

Example 2 ：双Y轴（Python）

import csv

+ ^8 v" H0 L" W5 E: `

import pandas as pd

import matplotlib.pyplot as plt

from datetime import datetime

( a2 ?, @; R0 b& m# j$ c' w& x

data=pd.read_csv(LOBO0010-2020112014010.tsv,sep=)

time=data[date [AST]]

) A6 r9 g$ A2 @

sal=data[salinity]

* u( h5 N. m9 c/ V. ]9 b, ]2 g

tem=data[temperature [C]]

; v, J! y/ m3 c$ R3 g$ \7 o3 ?/ b

print(sal)

DAT = []

for row in time:

; ^3 U1 s& h4 j( H

DAT.append(datetime.strptime(row,"%Y-%m-%d %H:%M:%S"))

#create figure

fig, ax =plt.subplots(1)

+ t; Q4 H# R% o" j& U( M

# Plot y1 vs x in blue on the left vertical axis.

plt.xlabel("Date [AST]")

* C( g# b Y z, l) j7 o; o

plt.ylabel("Temperature [C]", color="b")

' l" A( r6 ?" C; d4 E

plt.tick_params(axis="y", labelcolor="b")

- D/ X* N$ P- g6 v

plt.plot(DAT, tem, "b-", linewidth=1)

: t$ T& l, Z7 b' L$ [, a

plt.title("Temperature and Salinity from LOBO (Halifax, Canada)")

fig.autofmt_xdate(rotation=50)

! D/ ^# E. P5 z# [

# Plot y2 vs x in red on the right vertical axis.

plt.twinx()

plt.ylabel("Salinity", color="r")

6 J+ m) w- w7 o' g( c6 o

plt.tick_params(axis="y", labelcolor="r")

2 @. p7 k3 j! l' m# Z2 G( |6 @6 g$ b

plt.plot(DAT, sal, "r-", linewidth=1)

#To save your graph

plt.savefig(saltandtemp_V1.png ,bbox_inches=tight)

plt.show()

Example 3：拟合曲线（Python）

- e6 _1 |4 _+ w, D

import csv

import numpy as np

# {; q: S( C' A7 W W

import pandas as pd

from datetime import datetime

i4 z# V$ G0 m! G6 O& Y

import matplotlib.pyplot as plt

import scipy.signal as signal

data=pd.read_csv(LOBO0010-20201122130720.tsv,sep=)

% R# C- H% v% q

time=data[date [AST]]

temp=data[temperature [C]]

datestart = datetime.strptime(time[1],"%Y-%m-%d %H:%M:%S")

DATE,decday = [],[]

0 d- L, j8 X! c4 B$ U) h0 D8 Z

for row in time:

8 f4 c# r! \# T" W7 s# X3 R2 R

daterow = datetime.strptime(row,"%Y-%m-%d %H:%M:%S")

DATE.append(daterow)

decday.append((daterow-datestart).total_seconds()/(3600*24))

2 c! q: f4 H* x

# First, design the Buterworth filter

N = 2 # Filter order

* s' T6 w, |, u- X8 q6 W) A/ U

Wn = 0.01 # Cutoff frequency

B, A = signal.butter(N, Wn, output=ba)

# Second, apply the filter

& S! }0 Y2 ?: K8 X. B( Z# d

tempf = signal.filtfilt(B,A, temp)

# Make plots

, c3 l: z4 w O# f/ ~1 B

fig = plt.figure()

3 N Y5 M/ i7 d9 |3 e( e- A

ax1 = fig.add_subplot(211)

plt.plot(decday,temp, b-)

plt.plot(decday,tempf, r-,linewidth=2)

plt.ylabel("Temperature (oC)")

plt.legend([Original,Filtered])

' U6 H+ i/ K0 Q+ B) o+ u$ I. Q

plt.title("Temperature from LOBO (Halifax, Canada)")

ax1.axes.get_xaxis().set_visible(False)

ax1 = fig.add_subplot(212)

plt.plot(decday,temp-tempf, b-)

6 T5 _$ x* a1 u

plt.ylabel("Temperature (oC)")

# C" G8 @" t6 u; x

plt.xlabel("Date")

! d w* N# [' k. ?- u

plt.legend([Residuals])

8 a6 ^' ]' L% j' b& t6 b

plt.savefig(tem_signal_filtering_plot.png, bbox_inches=tight)

4 r! B' Z% H( I' X/ A0 q: B) c

plt.show()

/ g9 _) |8 d1 O. C

Example 4：三维地形（Python）

6 w3 l: U7 U. q

# This import registers the 3D projection

from mpl_toolkits.mplot3d import Axes3D

* a. k8 L0 C; F. W, o) b4 Z

from matplotlib import cbook

from matplotlib import cm

from matplotlib.colors import LightSource

/ s5 z2 u b+ m/ S2 ?/ j7 p

import matplotlib.pyplot as plt

import numpy as np

# G* c0 }4 g$ L% P: I6 X

filename = cbook.get_sample_data(jacksboro_fault_dem.npz, asfileobj=False)

+ A5 Y( u0 `4 ]

with np.load(filename) as dem:

7 b$ U7 n9 S+ S

z = dem[elevation]

nrows, ncols = z.shape

x = np.linspace(dem[xmin], dem[xmax], ncols)

0 x# i: Y& v+ ~- z. y% D

y = np.linspace(dem[ymin], dem[ymax], nrows)

x, y = np.meshgrid(x, y)

( h7 X- u8 \4 R6 Q, o6 a

region = np.s_[5:50, 5:50]

, F; j) g; j3 Z- ?! l

x, y, z = x[region], y[region], z[region]

" P8 \: ?. `2 q$ P2 {

fig, ax = plt.subplots(subplot_kw=dict(projection=3d))

ls = LightSource(270, 45)

rgb = ls.shade(z, cmap=cm.gist_earth, vert_exag=0.1, blend_mode=soft)

surf = ax.plot_surface(x, y, z, rstride=1, cstride=1, facecolors=rgb,

) n1 r# }+ g" E, W8 H9 N

linewidth=0, antialiased=False, shade=False)

8 e V% x' t; v# `

plt.savefig(example4.png,dpi=600, bbox_inches=tight)

plt.show()

) ^: A! x" g; X' z

Example 5：三维地形，包含投影（Python）

2 q7 p* i7 `( }7 @% c' s* R

Example 6：切片，多维数据同时展现（Python）

Example 7：SSH GIF 动图展现（Matlab）

w9 ~, v/ q) B& i# T

Example 8：Glider GIF 动图展现（Python）

8 |: ?6 [6 W0 F& K% y7 n2 Z

Example 9：涡度追踪 GIF 动图展现

* J4 i* {! W' c# m3 @3 L( t

5 q0 w" r$ o9 ^9 x2 o5 e

0 K! }3 Y- d+ D* t1 S