MesaData cannot handle one-entry history.data

[kailicao]

When history.data contains only one entry, MesaData cannot handle it, as the data(key) method returns an element instead of an array.

For instance, when I use the following code to read a one-entry history.data:
import mesa_reader as mr
h = mr.MesaData('history.data')

I get the following error:
Traceback (most recent call last):

File E:\MyResearch\MESA\1M_pre_ms_to_wd\LOGS_start\test.py:2 in
h = mr.MesaData('history.data')

File D:\Users\KailiCao\anaconda3\lib\site-packages\mesa_reader_init_.py:129 in init
self.read_data()

File D:\Users\KailiCao\anaconda3\lib\site-packages\mesa_reader_init_.py:177 in read_data
self.read_log_data()

File D:\Users\KailiCao\anaconda3\lib\site-packages\mesa_reader_init_.py:208 in read_log_data
self.remove_backups()

File D:\Users\KailiCao\anaconda3\lib\site-packages\mesa_reader_init_.py:673 in remove_backups
for i in range(len(self.data('model_number')) - 1):

TypeError: len() of unsized object

[ShaiAvr]

I am starting to learn MESA and I also encountered this bug. It’s weird that this issue exists for two years without a response. I can look into the code to try and fix this and make a pull request if the author doesn't have time to look at it.

[ShaiAvr]

I created a pull request #21 to fix this bug. Adding ndmin=1 to the np.genfromtxt call ensures we get a 1D array and not a scalar even with a single entry, so MesaData doesn't crash. Hopefully, this can be merged and published to pypi quickly.

[wmwolf]

Thanks, @ShaiAvr, #21 seems to solve the issue. I've merged it in and will include it in the next release once we resolve #20.