fft / inverse fft

fft_folder/preparation.py

@@ -0,0 +1,22 @@
+"""
+preparation.py
+
+fft
+inverse fft
+calculate frequencies
+"""
+import numpy as np
+import pandas as pd
+from datetime import datetime, timedelta
+
+def fft_powerspectrum(data):
+    """This function takes the function in and outputs
+    the powerspectrum"""
+    matrx = np.fft.fft(data.values)
+    return np.abs(matrx)[:len(matrx/2)]
+
+def fft_mag(data):
+    """this function is simalare to fft_powerspectrum only it does not cut the
+    matrix in half or take the absolut values of the variables"""
+    return np.fft.fft(data.values)
+

fft_folder/test_preparation.py

@@ -0,0 +1,72 @@
+"""
+test_preparation.py
+
+unit test for the functions fft, calc_freq, inv_fft
+
+"""
+import numpy as np
+import pandas as pd
+from datetime import datetime, timedelta
+import matplotlib.pyplot as plt
+
+from preparation import fft_powerspectrum, fft_mag
+#from thid import plot_rets
+
+"""
+t = np.linspace(0, 2, 2000, endpoint=False)
+f1, f2 = 50, 120
+data = pd.Series(np.sin(f1*t)+0.5*np.sin(f2*t),index=range(0,len(t)))
+# index_values = data.index.tolist()
+freq = calc_freq(data)
+print(freq)
+print(inv_fft(data))
+print(data)
+plot_rets(freq,fft(data))
+"""
+
+t = np.linspace(0, 2, 2000, endpoint=False)
+f1, f2 = 50, 120
+data = pd.Series(np.sin(2*np.pi*f1*t)+0.5*np.sin(2*np.pi*f2*t),index=range(0,len(t)))
+
+#trange = date_range(datetime.now(), datetime.now()+pd.timedelta(days=9),freq='d')
+#trange[i].timestamp()
+
+"""
+def test_fft(data):
+
+    magnitudes = fft(data)
+    assert len(magnitudes) == len(data)
+
+def test_calc_freq(data, f1, f2):            
+    freq = calc_freq(data)
+    lngth = len(data)/2
+    newf1 = f1 + lngth
+    newf2 = f2 + lngth
+    assert len(freq) == len(data)
+    assert freq[newf1] > 200
+    assert freq[newf2] > 200
+
+def test_inv_fft(data):
+    invdata = inv_fft(data)
+    assert len(invdata) == len(data)
+"""
+matr = fft_powerspectrum(data)
+
+print(matr)
+
+#freq = calc_freq(data)
+# plot_rets(freq,fft(data))
+
+#print(np.isclose(0, 0.0001, atol=0.001))
+
+trange = pd.date_range(datetime.now(), datetime.now()+timedelta(days=9),freq='d')
+data1 = pd.Series([1,2,3,4,5,6,7,8,9,10],index=trange)
+
+
+
+print(len(data1))
+
+print(data1.index[0].timestamp())
+
+
+print(data1.iloc[2] * 1)