Read xlsx #94
Read xlsx #94
Conversation
|
@jcrivenaes should we change the requirements as suggested here, or are there reasons to keep the previous requirements ? |
|
We should not enforce versions that are not yet in komodo stable. Pandas 1.1 is on its way in now, but xlrd might take another month or so. |
| ) | ||
|
|
||
| print(dgn.loc[0]["SENSCASE"]) |
There was a problem hiding this comment.
Is this a debugging statement?
| "Design matrix filename should be on Excel or csv format" | ||
| " and end with .xlsx or .csv" | ||
| "Design matrix should be on Excel or csv format" | ||
| " and filename should end with .xlsx or .csv" |
There was a problem hiding this comment.
"should -> must" is perhaps more to the point (.XLSX would even fail here, but we can live with that.)
| @@ -53,15 +53,17 @@ def summarize_design(filename, sheetname="DesignSheet01"): | |||
|
|
|||
| # Read design matrix and find realisation numbers for each sensitivity | |||
| if filename.endswith(".xlsx"): | |||
| dgn = pd.read_excel(filename, sheetname) | |||
| dgn = pd.read_excel(filename, sheetname, engine="openpyxl") | |||
| dgn.dropna(axis=0, how="all", inplace=True) | |||
There was a problem hiding this comment.
Is this due to things like "coloured cells"? If so, maybe nice to add a comment about that (to explain why the line of code is there, otherwise it could disappear in later refactorings)
|
Test pass on combinations of xlrd 1.2.0/2.0.1, and pandas versions 0.25 up to 1.2.0. Except Pandas 1.1.1, but this particular version is not in use. |
Changes pd.read_excel to use engine openpyxl since newer versions of xlrd does not support '.xlsx' files. Closes issue #88 . Requirements are updated to use pandas>1.1 and xlrd>2.0, but shold also work with previous requirements if we don't want to change these yet.