Converting to_data_frame for mne.io.Raw from 0.20.0 to 1.11

Support & Discussions
1

Hi,

I was working on a project where the MNE version is 0.20.0. I want to convert to 1.11 but I notice some issues.

In the 0.20.0 version, the method to_data_frame of mne.io.Raw has a parameter called scaling_time. However, in the 1.11 version, that method does not have that parameter anymore. How do I make sure that when I convert to the latest version, it behaves the same?

This is the original code, which is adapted from https://github.com/akaraspt/deepsleepnet/blob/master/prepare_physionet.py:

raw = mne.io.read_raw_edf(psg_fnames[i], preload=True, stim_channel=None)
raw_ch_df = raw.to_data_frame(scaling_time=100.0)[select_ch]
2

Hi,

The equivalent of scaling_time in 1.11 is time_format.

I’m having trouble accessing 0.20 documentation, but in 0.18 I see the default value of scaling_time is 1,000, which I’m guessing is a conversion from s to ms. CC @larsoner if I have this correct.

Default behaviour of the modern time_format is to return times as floats in s, but it does accept "ms" to return times as ints after rounding to the nearest ms.

However, in your case with scaling_time=100.0, I don’t think there is a 1:1 replacement. This would return times in tens of ms, right? If that’s really the unit you want, you could just get the dataframe with times as s using time_format=None, and then apply your scaling to the times column afterwards.

1 Like
3

This topic was automatically closed 7 days after the last reply. New replies are no longer allowed.