
import pandas,matplotlib,numpy, matplotlib.pyplot as pyplot


LMS_URL="https://www.uni-kiel.de/psychologie/golz/publications/2003a/LMS_SMJ_2.dat"
df=pandas.read_csv(LMS_URL, 
                   # sep="\s+", # 正規表現 \s:空白文字[\t\r\n\f], +:1以上の繰り返し
                   sep="\t",
                   header=None,
                   names=("wl","L","M","S"))


type(df)

pandas.core.frame.DataFrame


df[:5]


df.index, df.columns

(RangeIndex(start=0, stop=401, step=1),
 Index(['wl', 'L', 'M', 'S'], dtype='object'))


print(f"{df[0:3:1] = :}\n")

df[0:3:1] =     wl         L         M         S
0  380  0.000014  0.000006  0.001086
1  381  0.000018  0.000008  0.001453
2  382  0.000025  0.000010  0.001936


print(f" {df.L[0]= },\t\t {df['M'][0]= }")

 df.L[0]= 1.3813908e-05,		 df['M'][0]= 5.7853991e-06


print(f"{df.loc[0]= },\n\n{df.at[0,'S']= }")

df.loc[0]= wl    380.000000
L       0.000014
M       0.000006
S       0.001086
Name: 0, dtype: float64,

df.at[0,'S']= 0.0010858872


# ラベルを日本語にするために、fontを変更
font = {'family' : u'IPAGothic'}
matplotlib.rc('font', **font)
ax=df.plot.line("wl",["L","M","S"],
                title=u"錐体の応答(LMS刺激値)",
                xlabel="波長　[nm]",
                ylabel=u"応答",
                color=["r","g","b"],
                legend=True,
               )


type(ax)

matplotlib.axes._subplots.AxesSubplot


ax.set_xlabel(u"波長 [nm]", family=u'IPAGothic')
ax.set_ylabel(u"視細胞の応答", family=u'Noto Sans Japanese')
ax.set_title(u"錐体の応答 (LMS刺激値)", family=u'IPAGothic')
pyplot.draw()
ax.figure

<Figure size 432x288 with 0 Axes>


axes=df.plot('wl',['L','M','S'],
             kind='line', subplots=True,
             title="LMS 刺激値",
             xlabel="波長[nm]",
             ylabel=u"応答",
             color=["r","g","b"],
)


#logscale
axl=df.plot.line("wl",["L","M","S"],
                title="LMS 刺激値",
                xlabel="波長[nm]",
                ylabel=u"応答",
                figsize=(6,5.5),
                logy=True,
                 color=("r","g","b"),
                ylim=(1e-3,None)
)


ax.figure.savefig("./_images/LMS.png")


pyplot.imshow(pyplot.imread("./_images/LMS.png"));


#　光の3原色(RGB)
lambda_R=700 # in nm
lambda_G=546.1
lambda_B=435.8


ymin,ymax=ax.get_ylim()
ax.arrow(lambda_R,-0.1, 0,0, 
          width=0.1,head_width=10, head_length=0.05,
          color="red",head_starts_at_zero=False)
ax.arrow(lambda_G,-0.1, 0,0 ,
          width=0.1,head_width=10, head_length=0.05,
         color="green",head_starts_at_zero=False)
ax.arrow(lambda_B,-0.1, 0,0,
          width=0.1,head_width=10, head_length=0.05,
         color="blue",head_starts_at_zero=False)
ax.figure


ax.annotate("R",(lambda_R, -0.1), 
            (0,-24),textcoords="offset points",color="red",fontsize=16,
            arrowprops=dict(facecolor='red', shrink=0.05),
            horizontalalignment='right', verticalalignment='top'
           )
ax.annotate("G",(lambda_G, -0.1),
            (0,-24),textcoords="offset points",color="green",fontsize=16,
            arrowprops=dict(facecolor='green', shrink=0.05),
            horizontalalignment='right', verticalalignment='top'
           )
ax.annotate("B",(lambda_B, -0.1),
            (0,-24), textcoords="offset points",color="blue",fontsize=16,
            arrowprops=dict(facecolor='blue', shrink=0.05),
            horizontalalignment='right', verticalalignment='top'
           )
ax.figure


na= df.values; # also na=df.to_numpy()
type(na)

numpy.ndarray


na.dtype

dtype('float64')


na.shape

(401, 4)


na[:2,:] #最初の二行を表示

array([[3.8000000e+02, 1.3813908e-05, 5.7853991e-06, 1.0858872e-03],
       [3.8100000e+02, 1.8467576e-05, 7.7568882e-06, 1.4526153e-03]])


na.shape=(401,2,2)
na[:2,:]

array([[[3.8000000e+02, 1.3813908e-05],
        [5.7853991e-06, 1.0858872e-03]],

       [[3.8100000e+02, 1.8467576e-05],
        [7.7568882e-06, 1.4526153e-03]]])


na.shape=(401,4) # 元に戻す。
print(f"{na[-1,0] =:}\n") #最後の行の最初の列の要素
print(f"{na[0] =:}\n") # 最初の一行
print(f"{na[0:2] =:}\n") # 最初の2行
print(f"{na[0:2, :] =:}\n") # 最初の2行
print(f"{na[0:2,...] =:}\n") # 最初の2行

na[-1,0] =780.0

na[0] =[3.8000000e+02 1.3813908e-05 5.7853991e-06 1.0858872e-03]

na[0:2] =[[3.8000000e+02 1.3813908e-05 5.7853991e-06 1.0858872e-03]
 [3.8100000e+02 1.8467576e-05 7.7568882e-06 1.4526153e-03]]

na[0:2, :] =[[3.8000000e+02 1.3813908e-05 5.7853991e-06 1.0858872e-03]
 [3.8100000e+02 1.8467576e-05 7.7568882e-06 1.4526153e-03]]

na[0:2,...] =[[3.8000000e+02 1.3813908e-05 5.7853991e-06 1.0858872e-03]
 [3.8100000e+02 1.8467576e-05 7.7568882e-06 1.4526153e-03]]


na[(0,200,400),...] #指定した行(0,200,400)をとりだす。

array([[3.8000000e+02, 1.3813908e-05, 5.7853991e-06, 1.0858872e-03],
       [5.8000000e+02, 6.7649208e-01, 2.3602925e-01, 5.2250678e-04],
       [7.8000000e+02, 1.4279820e-05, 6.3144453e-07, 1.3166252e-08]])


na[::200,...] # 200行毎にとりだす。ellipse(...)を使った。

array([[3.8000000e+02, 1.3813908e-05, 5.7853991e-06, 1.0858872e-03],
       [5.8000000e+02, 6.7649208e-01, 2.3602925e-01, 5.2250678e-04],
       [7.8000000e+02, 1.4279820e-05, 6.3144453e-07, 1.3166252e-08]])


na[::200,:] # 200行毎にとりだす。スライス(: )

array([[3.8000000e+02, 1.3813908e-05, 5.7853991e-06, 1.0858872e-03],
       [5.8000000e+02, 6.7649208e-01, 2.3602925e-01, 5.2250678e-04],
       [7.8000000e+02, 1.4279820e-05, 6.3144453e-07, 1.3166252e-08]])


na[slice(0,401,200),...] # 200行毎にとりだす。

array([[3.8000000e+02, 1.3813908e-05, 5.7853991e-06, 1.0858872e-03],
       [5.8000000e+02, 6.7649208e-01, 2.3602925e-01, 5.2250678e-04],
       [7.8000000e+02, 1.4279820e-05, 6.3144453e-07, 1.3166252e-08]])


import matplotlib.pyplot as pyplot
pyplot.ion()
pyplot.plot(na[:,0],na[:,1],"r", label="L")
pyplot.legend()
#pyplot.draw()
#pyplot.show()

<matplotlib.legend.Legend at 0x12a96f640>


lines=pyplot.plot(na[:,0],na[:,1],"r",
                  na[:,0],na[:,2],"g",
                  na[:,0],na[:,3],"b")
pyplot.legend(["L","M","S"])
pyplot.xlabel("波長[nm]")
pyplot.ylabel("応答")

Text(0, 0.5, '応答')


fig=pyplot.figure(constrained_layout=False,figsize=(8,4)) # inches
ax=fig.add_subplot(1,1,1)
ax.plot(na[:,0],na[:,1],"r",label="L")
ax.plot(na[:,0],na[:,2],"g",label="M")
ax.plot(na[:,0],na[:,3],"b",label="S")
ax.set_title(u"三錐体(LMS)の応答")
ax.set_xlabel("波長[nm]", family="IPAGothic")
ax.set_ylabel(u"応答")
ax.set_xlim(None,None)
ax.legend()

<matplotlib.legend.Legend at 0x12aa9fd30>


fig=pyplot.figure(constrained_layout=True,figsize=(8,4))
axes=fig.subplots(3,1)
fig.suptitle("LMS 応答関数（個別)")
for i, ax, c, lbl in zip(range(1,4), axes, ["r","g","b"], ["L","M","S"]):
    ax.plot(na[:,0],na[:,i],c,label=lbl)
    ax.legend()
    ax.set_ylabel("応答")
axes[-1].set_xlabel("波長[nm]")
pyplot.draw()


fig=pyplot.figure(figsize=(9,4), constrained_layout=True)
fig.suptitle("LMS curve in 3D");
zrot=2
axL=fig.add_subplot(1,2,1,projection='3d',
                   azim=-30-zrot,elev=30,proj_type='persp')
axR =fig.add_subplot(1,2,2,projection='3d',
                     azim=-30+zrot,elev=30,proj_type='persp',
                     sharez=axL,sharex=axL,sharey=axL)
def plotLMS3D(ax,na):
    ax.set_xlabel('L')
    ax.set_ylabel('M')
    ax.set_zlabel('S')
    L,M,S =na[:,1], na[:,2], na[:,3]
    ax.plot(L,M,S,label="LMS")
plotLMS3D(axL,na);
plotLMS3D(axR,na);
axL.set_title("Left");
axR.set_title("Right");


fig=pyplot.figure(figsize=(9,4), constrained_layout=True)
fig.suptitle("lms curve in 3D");
zrot=2
axL=fig.add_subplot(1,2,1,projection='3d',
                   azim=-30-zrot,elev=30,proj_type='persp')
axR =fig.add_subplot(1,2,2,projection='3d',
                     azim=-30+zrot,elev=30,proj_type='persp',
                     sharez=axL,sharex=axL,sharey=axL)
def plotLMS3D(ax,na):
    ax.set_xlabel('l') ;   ax.set_ylabel('m');    ax.set_zlabel('s')
    L,M,S =na[:,1], na[:,2], na[:,3]
    Y=L+M+S
    L,M,S=L/Y,M/Y,S/Y
    ax.plot(L,M,S,label="lms")
plotLMS3D(axL,na);
plotLMS3D(axR,na);
axL.set_title("Left");
axR.set_title("Right");


%matplotlib --list

Available matplotlib backends: ['tk', 'gtk', 'gtk3', 'gtk4', 'wx', 'qt4', 'qt5', 'qt6', 'qt', 'osx', 'nbagg', 'notebook', 'agg', 'svg', 'pdf', 'ps', 'inline', 'ipympl', 'widget']


def lininterp(v):
    w=[e[0] for e in v]
    s=[e[1] for e in v]
    def interp(x,_w=w,_s=s):
        i=j=None
        if int(x) in _w:
            i=_w.index(int(x))
        if int(x)+1 in _w:
            j=_w.index(int(x)+1)
        if i and j:
            ret=(_s[i]*(_w[j]-x)+_s[j]*(x-_w[i]))/(_w[j]-_w[i])
            return ret
        elif i:
            return _s[i]
        elif j:
            return _s[j]
        else:
            return 0
    return interp


lambda_R, lambda_G, lambda_B =700, 546.1,435.8
L_tbl ,M_tbl, S_tbl =[ na[:,[0,i]] for i in range(1,4)]
L,M,S=lininterp(L_tbl),lininterp(M_tbl),lininterp(S_tbl)


mat_RGB_to_LMS=numpy.matrix((
    (L(lambda_R), L(lambda_G), L(lambda_B)),
    (M(lambda_R), M(lambda_G), M(lambda_B)),
    (S(lambda_R), S(lambda_G), S(lambda_B))
))
mat_LMS_to_RGB=mat_RGB_to_LMS**(-1)

print(f"{mat_RGB_to_LMS = !r}\n\n{mat_LMS_to_RGB = !r}")

mat_RGB_to_LMS = matrix([[4.02808400e-03, 6.51276398e-01, 2.14291164e-02],
        [1.41605110e-04, 3.63393490e-01, 1.85730100e-02],
        [4.40129960e-07, 6.78473818e-03, 1.85053618e+00]])

mat_LMS_to_RGB = matrix([[ 2.64950902e+02, -4.74878594e+02,  1.69802741e+00],
        [-1.03260691e-01,  2.93743062e+00, -2.82859332e-02],
        [ 3.15575522e-04, -1.06567435e-02,  5.40487225e-01]])


print(f"{mat_LMS_to_RGB@mat_RGB_to_LMS = !r}\n\n{mat_RGB_to_LMS@mat_LMS_to_RGB = !r}")

mat_LMS_to_RGB@mat_RGB_to_LMS = matrix([[ 1.00000000e+00,  4.93528829e-14,  1.77635684e-15],
        [-1.77347523e-20,  1.00000000e+00, -6.93889390e-18],
        [-7.94093388e-23,  4.33680869e-19,  1.00000000e+00]])

mat_RGB_to_LMS@mat_LMS_to_RGB = matrix([[ 1.00000000e+00,  9.50845305e-17,  0.00000000e+00],
        [ 7.46913653e-18,  1.00000000e+00, -1.73472348e-18],
        [ 0.00000000e+00,  0.00000000e+00,  1.00000000e+00]])


import numpy.linalg as linalg
LMS=na[:,1:]
print(f"{LMS.shape = }")
#LMS.shape=(LMS.shape[0],3,1)
print(LMS[0])
RGB=LMS@(mat_LMS_to_RGB.transpose())
print(RGB[0],RGB.shape)

LMS.shape = (401, 3)
[1.3813908e-05 5.7853991e-06 1.0858872e-03]
[[ 2.75651142e-03 -1.51475581e-05  5.86850866e-04]] (401, 3)


fig=pyplot.figure()
ax=fig.add_subplot(1,1,1)
for i,c in enumerate(["r","g","b"]):
    ax.plot(na[:,0],RGB[:,i],c,label=c.upper())
ax.legend()

<matplotlib.legend.Legend at 0x12ad2e7c0>


fig=pyplot.figure()
ax=fig.add_subplot()
RGB_scale=[1/max(RGB[:,i]) for i in range(3)] 
for i,c in enumerate(["r","g","b"]):
    ax.plot(na[:,0], RGB[:,i]*RGB_scale[i], c,label=c.upper())
ax.legend()

<matplotlib.legend.Legend at 0x12ad62c40>


mat_wRGB_to_wLMS=numpy.matrix((
    (1, 0, 0, 0),
    (0, L(lambda_R), L(lambda_G), L(lambda_B)),
    (0, M(lambda_R), M(lambda_G), M(lambda_B)),
    (0, S(lambda_R), S(lambda_G), S(lambda_B))
))
mat_wLMS_to_wRGB=mat_wRGB_to_wLMS**(-1)

print(f"{mat_wRGB_to_wLMS = !r}\n\n{mat_wLMS_to_wRGB = !r}\n")
print(f"{mat_wLMS_to_wRGB@mat_wRGB_to_wLMS = !r}\n\n{mat_wRGB_to_wLMS@mat_wLMS_to_wRGB = !r}")

mat_wRGB_to_wLMS = matrix([[1.00000000e+00, 0.00000000e+00, 0.00000000e+00, 0.00000000e+00],
        [0.00000000e+00, 4.02808400e-03, 6.51276398e-01, 2.14291164e-02],
        [0.00000000e+00, 1.41605110e-04, 3.63393490e-01, 1.85730100e-02],
        [0.00000000e+00, 4.40129960e-07, 6.78473818e-03, 1.85053618e+00]])

mat_wLMS_to_wRGB = matrix([[ 1.00000000e+00,  0.00000000e+00,  0.00000000e+00,
          0.00000000e+00],
        [ 0.00000000e+00,  2.64950902e+02, -4.74878594e+02,
          1.69802741e+00],
        [ 0.00000000e+00, -1.03260691e-01,  2.93743062e+00,
         -2.82859332e-02],
        [ 0.00000000e+00,  3.15575522e-04, -1.06567435e-02,
          5.40487225e-01]])

mat_wLMS_to_wRGB@mat_wRGB_to_wLMS = matrix([[ 1.00000000e+00,  0.00000000e+00,  0.00000000e+00,
          0.00000000e+00],
        [ 0.00000000e+00,  1.00000000e+00,  2.09311735e-14,
          8.88178420e-16],
        [ 0.00000000e+00, -1.77347523e-20,  1.00000000e+00,
         -6.93889390e-18],
        [ 0.00000000e+00, -7.94093388e-23,  4.33680869e-19,
          1.00000000e+00]])

mat_wRGB_to_wLMS@mat_wLMS_to_wRGB = matrix([[ 1.00000000e+00,  0.00000000e+00,  0.00000000e+00,
          0.00000000e+00],
        [ 0.00000000e+00,  1.00000000e+00,  9.50845305e-17,
          0.00000000e+00],
        [ 0.00000000e+00,  5.30242625e-19,  1.00000000e+00,
         -1.73472348e-18],
        [ 0.00000000e+00,  0.00000000e+00,  0.00000000e+00,
          1.00000000e+00]])


av_phosphors=pandas.read_csv(
    "https://www.uni-kiel.de/psychologie/golz/publications/2003a/AveragePhosphors.dat",
    sep="\t",
    header=None,
    names=("wl","R","G","B")
)


av_phosphors.plot.line("wl",["R","G","B"],color=["red","green","blue"])

<AxesSubplot:xlabel='wl'>


scale=numpy.diag( [1/av_phosphors.values[:,i].sum() for i in range(1,4)])
NRGB=av_phosphors.values[:,1:]@scale
RGB=av_phosphors.values[:,1:]
fig=pyplot.figure()
fig.suptitle("ディスプレィ蛍光体の平均発光スペクトラム\n正規化済み")
ax=fig.add_subplot(1,1,1)
ax.plot(av_phosphors["wl"],NRGB[:,0],"r",label="R")
ax.plot(av_phosphors["wl"],NRGB[:,1],"g",label="G")
ax.plot(av_phosphors["wl"],NRGB[:,2],"b",label="B")
ax.legend()

<matplotlib.legend.Legend at 0x12a92fc10>


import scipy.io, io, urllib.request
monitors=scipy.io.loadmat(
    io.BytesIO(
        urllib.request.urlopen(
            "https://www.uni-kiel.de/psychologie/golz/publications/2003a/Phosphors.mat"
        ).read())
)["phosphors"]
monitors.shape

(401, 3, 15)


fig=pyplot.figure(figsize=(12,6))
fig.suptitle("各種モニターでの蛍光体 発光特性", family=u'IPAMincho',fontsize=24)
axes=fig.subplots(3,5)
for i,ax in enumerate(axes.flatten()):
    ax.plot(na[:,0], monitors[:,0,i],color="red",label="R")
    ax.plot(na[:,0], monitors[:,1,i],color="green",label="G")
    ax.plot(na[:,0], monitors[:,2,i],color="blue",label="B")
pyplot.draw()

	wl	L	M	S
0	380	0.000014	0.000006	0.001086
1	381	0.000018	0.000008	0.001453
2	382	0.000025	0.000010	0.001936
3	383	0.000033	0.000014	0.002568
4	384	0.000043	0.000018	0.003390

Python 入門講座 第7回: タイトル募集中¶

使用するデータについて¶

CSVファイルを読み込む: pandasを使ってみる。¶

Pandas dataframe¶

データフレームは２次元の表データフレームはこの様に、行および列にそれぞれラベル(indexおよび column names)を持つ2次元の表として表現されます。¶

データフレームの成分の取り出し¶

データフレームからグラフを作成¶

プロットオプション：サブプロット¶

プロットオプション: 対数表示¶

グラフの保存¶

保存したグラフの表示¶

3原色表示¶

マーカーの追加¶

アノテーションの追加¶

matplotlibとnumpy ndarray¶

matplotlib¶

ndarray in numpy モジュール¶

データフレームから ndarrayへ¶

shape¶

shapeの変更¶

ndarrayの要素の取り出し¶

pyplot モジュールによるグラフの作成¶

複数のデータを含むグラフ¶

タイトル ラベルの追加¶

サブプロット¶

3D プロット¶

pyplot.draw(), pyplot.show(), pyplot.ion(),pyplot.ioffについて¶

non-interactiveモード（デフォールト）¶

pyplot.draw()¶

pyplot.show()¶

interactiveモード¶

jupyterlab Notebookの%matplotlib マジックについて¶

3原色の色覚曲線¶

線形補間関数を作る。¶

RGBからLMSへの変換¶

等色関数をプロットする。¶

蛍光体のスペクトラム¶

この発光体のスペクトラムをグラフに表示してみましょう。¶

スペクトルの正規化¶

.matファイル(MATLABのデータファイル形式)の読み込み¶

今日のまとめ¶

Python 入門講座　第7回: タイトル募集中¶

　タイトル　ラベルの追加¶

`pyplot.draw()`, `pyplot.show()`, `pyplot.ion()`,`pyplot.ioff`について¶

`pyplot.draw()`¶

`pyplot.show()`¶

jupyterlab Notebookの`%matplotlib` マジックについて¶

　蛍光体のスペクトラム¶