Gert Hulselmans

To me it looks like your expression matrix does not contain TABs but spaces as separators. ```bash awk -v OFS='\t' '$1 = $1' /d/pyscenic/data/FB6_countsMatrix.tsv > /d/pyscenic/data/FB6_countsMatrix.with_tabs.tsv ```

Login you your server with `ssh -X`. Or don't use `import tkinter as tk` and ` matplotlib.use('TkAgg')` ```python import matplotlib matplotlib.use('Agg') import matplotlib.pyplot as plt ```

Preferably it would just rely on the first 4 magic bytes to determine if it is a bigwig or bigbed file and not look at the extension at all: https://github.com/jackh726/bigtools/blame/master/bigtools/src/bbi/bbiread.rs#L657-L664...

Compiling without llvm probably makes it a lot smaller. My frawk binary is 9.4M stripped and 17M unstripped. I rarely see `mawk` being much faster than `frawk` (except for https://github.com/ezrosent/frawk/issues/98)....

``` # Without LLVM, but with other recommended defaults $ cargo +nightly install --path . --no-default-features --features use_jemalloc,allow_avx2,unstable ```

The awk scripts that don't work, need to be reformatted (`frawk` parser does not accept all valid awk script yet): ```diff - mawk '{ ++count[$0] } - END { -...

@9ao9ai9ar For what it is worth, in my branch (https://github.com/ghuls/frawk/tree/updated_frawk) more `awk` scripts can be parsed without reformatting: https://github.com/ezrosent/frawk/pull/123 Only a tiny change is still needed, as `awk` allows executing...

Run with less number of workers to reduce memory pressure (and avoiding killing of your processes).

Did you ran out of disk space, by any chance?

The implementation of export_pseudobulk was rewritten to be faster, hence the slightly different logging lines.