Plotting histogram in years like BCE, CE

I want to plot a histogram of frequency of eruptions based on the years.

However, the data provided is not so simple. It has year in the form of BCE and CE, along with "unknowns". I want to get rid of the unknowns and count the number of eruptions. However, I cannot read the data as an integer/etc as there are strings in it. "500 CE" and "205 BCE" for example. Is there a way to plot the histogram accounting for such data? I have attached a photo of the data sample in the link.

So far, what I have is

 % Locating the index for unknowns
 ridunknowns = find(strcmp('Unknown',eruptyear)); 
% Changing the values of Unknowns to NaN in the vector

 eruptyear([ridunknowns]) = {NaN};

Answer 1

[Steven Lord]

Define the format for your date data.

fmt = 'yyyy G';

Convert your data into a datetime array. "Unknown" dates become NaT.

S = {'3300 BCE'; '4040 BCE'; 'Unknown'; '1944 CE'};
DT = datetime(S, 'InputFormat', fmt, 'Format', fmt)

Generate 1000 year wide bins.

startBin = datetime('5000 BCE', 'InputFormat', fmt, 'Format', fmt);
endBin = datetime('2001 CE', 'InputFormat', fmt, 'Format', fmt);
E = startBin:calyears(1000):endBin

Create a histogram using the selected edges.

histogram(DT, E)

If instead you want century-wide bins you can specify that as the BinMethod.

histogram(DT, 'BinMethod', 'century')

[Danny]

Can one use only the positive or negative sign instead of BCE/CE ?