mikeio.generic.concat first/last not working

[FransDHI]

Describe the bug
Using mikeio.generic.concat on a list of .dfs0 to concat the files -> always keeping the last regardless of the keep='last' / 'first' flag.

To Reproduce
dfs0_fls = glob(os.path.join(outpath,'*.dfs0'))
mikeio.generic.concat(infilenames=dfs0_fls,outfilename=os.path.join(outpath,'HD_Analysis_Loc_1979_2024_Keep_First.dfs0'),keep='first')
mikeio.generic.concat(infilenames=dfs0_fls,outfilename=os.path.join(outpath,'HD_Analysis_Loc_1979_2024_Keep_Last.dfs0'),keep='last')

Both produce the same result at the "join" time steps -> they both keep 'last'

Problem is in def _concat_time of (https://github.com/DHI/mikeio/tree/main/src/mikeio/dataset/_dataset.py)
lines 1152 - 1157

for j in range(ds.n_items):
if keep == "last":
newdata[j][idx1] = ds[j].to_numpy()
newdata[j][idx2] = other[j].to_numpy()
else:
newdata[j][idx2] = other[j].to_numpy()
newdata[j][idx1] = ds[j].to_numpy()

System information:

• Python version
• MIKE IO version: latest

[ecomodeller]

Hi @FransDHI,

Please note this in glob.glob
Note The pathnames are returned in no particular order. If you need a specific order, sort the results.

Please check if this fixes the problem (adjust the sorting according to your naming convention).

dfs0_fls = sorted(glob(os.path.join(outpath, '*.dfs0')))

[FransDHI]

No. It is not glob. I tested that with only 2 dfs0s. It is the the if / else statement shown above. There is no difference in the code following the if and else. It does the same thing. so the flag for first / last will always execute as if last. So the index will always be set to the first index of last. See below:

for j in range(ds.n_items):
if keep == "last":
newdata[j][idx1] = ds[j].to_numpy() ### 1. This Line is the same as Line 4. in the else statement
newdata[j][idx2] = other[j].to_numpy() ### 2. This Line is the same as Line 3. in the else statement
else:
newdata[j][idx2] = other[j].to_numpy() ### 3.
newdata[j][idx1] = ds[j].to_numpy() ### 4.

[ecomodeller]

The difference between the branches is the order. The indexes idx1 and idx2 overlap (which is the reason why have the 'last' and 'first' options).

[FransDHI]

see above code re-ordered for clarity:

for j in range(ds.n_items):
if keep == "last":
newdata[j][idx2] = other[j].to_numpy() ### 2. This Line is the same as Line 3. in the else statement
newdata[j][idx1] = ds[j].to_numpy() ### 1. This Line is the same as Line 4. in the else statement
else:
newdata[j][idx2] = other[j].to_numpy() ### 3.
newdata[j][idx1] = ds[j].to_numpy() ### 4.

What is the difference between the code following the if and else statements ?