When I read from an mdf (.mf4) file with asammdf (Python 3.7), as soon as the sample count crosses a barrier, the timestamps start counting up from close to 0 again, like an overflow :for 3 files exactly at 29045 lines, one file at 27234 lines for some reason. This means that when I use methods like resample or to_dataframe that the interpolation that occurs during these methods royally screws up and I get botched data.
I haven't found anything of the sort in the documentation, and there's hardly any resources besides the doc. I thought this might be related to chunk size or memory allocation, but I can't figure out what to do differently or why exactly this happens.
Right now I read it via the standard method
mdf = MDF(file)
I wrote up a small script to read in a file and plot the difference between the converted (interpolated) data and the original datapoints to showcase what I mean:
import tkinter as tk
from tkinter import filedialog as fd
from asammdf import MDF
from matplotlib import pyplot as plt
typeStr = '*.mf4'
root = tk.Tk()
root.wm_attributes('-topmost', 1)
root.withdraw()
files = fd.askopenfilenames(parent=root,filetypes=[("Measurement MF4 file",typeStr)])
for file in files:
mdf = MDF(file)
# conversion to pandas
data = mdf.to_dataframe()
data['Time [s]'] = list(data.index)
columns = data.columns.tolist()
columns.remove('Time [s]')
columns.insert(0,'Time [s]')
data = data[columns]
plt.plot(data['Time [s]'], data[columns[1]],label="converted")
# original data
chData = mdf.get(columns[1])
plt.plot(chData.timestamps, chData.samples, label="original")
plt.legend()
plt.show()
an example file can be accessed here:
an example of a plot:
