The ASAP CRN convention is to encode NULL values in metadata as "NA". We have wrappers in crn-utils (read_meta_table) to handle this, but the analysis scripts do NOT.
When you use pd.read_csv(), Pandas automatically interprets certain strings as missing values (NaN). These include:
'' (empty string), 'NA, 'N/A, '#N/A, '#N/A N/A, 'NaN, 'nan, 'null, 'NULL, '-1.#IND, '-1.#QNAN, '1.#IND, '1.#QNAN, ', and 'None.
To control this behavior:
na_values:
You can specify additional strings to be treated as NaN using the na_values parameter.
keep_default_na:
If you want to disable the default NaN recognition, set keep_default_na=False.
The ASAP CRN convention is to encode NULL values in metadata as "NA". We have wrappers in
crn-utils(read_meta_table) to handle this, but the analysis scripts do NOT.When you use pd.read_csv(), Pandas automatically interprets certain strings as missing values (NaN). These include: '' (empty string), 'NA, 'N/A, '#N/A, '#N/A N/A, 'NaN, 'nan, 'null, 'NULL, '-1.#IND, '-1.#QNAN, '1.#IND, '1.#QNAN, ', and 'None.
To control this behavior:
na_values:
You can specify additional strings to be treated as NaN using the na_values parameter.
keep_default_na:
If you want to disable the default NaN recognition, set keep_default_na=False.
See also: https://github.com/ASAP-CRN/pmdbs-bulk-rnaseq-wf/issues/9#issue-2657377065