ドル円の基本要約量
import pandas as pd from pandas.plotting import autocorrelation_plot import numpy as np from scipy import stats from matplotlib import pylab as plt import seaborn as sns sns.set() import statsmodels.api as sm import heapq as hp import math fx_data = pd.read_csv("/Users/ユーザー名/Documents/FX/foreign_exchange_historical_data/USDJPY/USDJPY_DAY.csv")#データの読み込み #pandasの計算は遅いので、全てnumpy配列に変換 opening = np.array(fx_data["opening"])#始値 high = np.array(fx_data["high"])#高値 low = np.array(fx_data["low"])#低値 closing = np.array(fx_data["closing"])#終値 #平均 opening_mean = np.mean(opening) high_mean = np.mean(high) low_mean = np.mean(low) closing_mean = np.mean(closing) #標準偏差 opening_sd = np.std(opening) high_sd = np.std(high) low_sd = np.std(low) closing_sd = np.std(closing)
結果は以下。
#平均 opening, high, low, closing 102.00780350978135, 102.4274237629459, 101.5360411392405, 102.01047871116225 #標準偏差 opening, high, low, closing 13.128446697845026, 13.144156607787117, 13.101283871625002, 13.132973566366692
- その他
#最大値、最小値 max(high), min(low) 125.86, 75.57 #1日の変動の最大値、最小値 max(high-low), min(high-low) 7.929999999999993, 0.03999999999999204 #1日の変動の平均値、標準偏差 np.mean(abs(closing-opening)), np.std(abs(closing-opening)) 0.4465123705408514, 0.4547812358889767 #上昇と下降の割合 co = closing - opening len(co[co >0])/len(co) 0.49108170310701954 #有意性検定を行うほどの差は考えにくい
- 簡単にやる時は、次のようにする。
fx_data.describe()
opening high low closing
count 3249.000000 3249.000000 3249.000000 3249.000000
mean 101.057579 101.473152 100.599625 101.064289
std 12.969594 12.987922 12.946784 12.976581
min 75.760000 75.980000 75.570000 75.680000
25% 90.900000 91.300000 90.250000 90.840000
50% 104.210000 104.640000 103.690000 104.190000
75% 110.620000 110.911000 110.240000 110.630000
max 125.660000 125.860000 124.540000 125.550000
- 解析期間(2007年4月2日から2020年8月15日）
