selimfirat

Issues somebody may already be working on are listed last. List everything by date

• Test Python 3.14 and current scikit-learn in CI, and fix the two failures they surface

  Open

  ⋯Report a problem with this issue

  • Is something wrong?

  • Already taken
  • Not beginner-friendly
  • Stale or invalid
  • Something else…
  • Send Cancel
  enhancement help wanted

  Difficulty 4/5 3-5 days Newbie friendliness 68/100

  selimfirat/pysad#160 · Opened Sep 28, 2026

  Maintainers usually reply within 1 day

• Bug in Inqmad: fitted model can't be pickled

  Open

  ⋯Report a problem with this issue

  • Is something wrong?

  • Already taken
  • Not beginner-friendly
  • Stale or invalid
  • Something else…
  • Send Cancel
  bug help wanted

  Difficulty 3/5 1-2 days Newbie friendliness 74/100

  selimfirat/pysad#159 · Opened Sep 28, 2026

  Maintainers usually reply within 1 day

• Implement the concept drift check of IForestASD (Algorithm 2 in Ding and Fei, 2013)

  Open

  ⋯Report a problem with this issue

  • Is something wrong?

  • Already taken
  • Not beginner-friendly
  • Stale or invalid
  • Something else…
  • Send Cancel
  enhancement help wanted

  Difficulty 4/5 3-5 days Newbie friendliness 68/100

  selimfirat/pysad#158 · Opened Sep 28, 2026

  Maintainers usually reply within 1 day

• Examples and the docs quick start fail after pip install: let Data download datasets on demand

  Open

  ⋯Report a problem with this issue

  • Is something wrong?

  • Already taken
  • Not beginner-friendly
  • Stale or invalid
  • Something else…
  • Send Cancel
  enhancement help wanted

  Difficulty 4/5 3-5 days Newbie friendliness 48/100

  selimfirat/pysad#157 · 1 comment · Opened Sep 28, 2026

  Maintainers usually reply within 1 day

• Bug in Inqmad: rho accumulates in float32, so updates lose precision on long streams

  Possibly taken @selimfirat claimed this 1 day ago. Open

  ⋯Report a problem with this issue

  • Is something wrong?

  • Already taken
  • Not beginner-friendly
  • Stale or invalid
  • Something else…
  • Send Cancel
  bug

  selimfirat/pysad#175 · 1 assignee · Opened Sep 28, 2026

  Maintainers usually reply within 1 day

• Inqmad: the class-level jit caches keep every instance's feature map and executables alive

  Possibly taken @selimfirat claimed this 1 day ago. Open

  ⋯Report a problem with this issue

  • Is something wrong?

  • Already taken
  • Not beginner-friendly
  • Stale or invalid
  • Something else…
  • Send Cancel
  bug

  selimfirat/pysad#173 · 1 assignee · Opened Sep 28, 2026

  Maintainers usually reply within 1 day

• Bug in Inqmad: no input validation beyond the single-instance check (feature count, 0-d input, batch_size)

  Possibly taken @selimfirat claimed this 1 day ago. Open

  ⋯Report a problem with this issue

  • Is something wrong?

  • Already taken
  • Not beginner-friendly
  • Stale or invalid
  • Something else…
  • Send Cancel
  bug

  selimfirat/pysad#172 · 1 assignee · Opened Sep 28, 2026

  Maintainers usually reply within 1 day

• Bug in Inqmad: QFeatureMap_rff.update_rff has no effect once the feature map has been called

  Possibly taken @selimfirat claimed this 1 day ago. Open

  ⋯Report a problem with this issue

  • Is something wrong?

  • Already taken
  • Not beginner-friendly
  • Stale or invalid
  • Something else…
  • Send Cancel
  bug

  selimfirat/pysad#171 · 1 assignee · Opened Sep 28, 2026

  Maintainers usually reply within 1 day

• RelativeEntropy: the rarity threshold c_th is fixed at 1, which makes the per-hypothesis window counts dead code

  Possibly taken @selimfirat claimed this 1 day ago. Open

  ⋯Report a problem with this issue

  • Is something wrong?

  • Already taken
  • Not beginner-friendly
  • Stale or invalid
  • Something else…
  • Send Cancel
  bug

  selimfirat/pysad#166 · 1 comment · 1 assignee · Opened Sep 28, 2026

  Maintainers usually reply within 1 day

• RelativeEntropy keeps every value in util, so memory grows with the stream although only the last window_size values are used

  Possibly taken @selimfirat claimed this 1 day ago. Open

  ⋯Report a problem with this issue

  • Is something wrong?

  • Already taken
  • Not beginner-friendly
  • Stale or invalid
  • Something else…
  • Send Cancel
  bug

  selimfirat/pysad#165 · 1 assignee · Opened Sep 28, 2026

  Maintainers usually reply within 1 day

• Bug in RelativeEntropy: num_bins, min_val and max_val are not validated (num_bins=1 gives T = nan, min_val > max_val gives a negative bucket width)

  Possibly taken @selimfirat claimed this 1 day ago. Open

  ⋯Report a problem with this issue

  • Is something wrong?

  • Already taken
  • Not beginner-friendly
  • Stale or invalid
  • Something else…
  • Send Cancel
  bug

  selimfirat/pysad#164 · 1 assignee · Opened Sep 28, 2026

  Maintainers usually reply within 1 day

• Bug in RSHash: an integer or zero decay fails at construction with an unrelated NumPy error

  Possibly taken @selimfirat claimed this 1 day ago. Open

  ⋯Report a problem with this issue

  • Is something wrong?

  • Already taken
  • Not beginner-friendly
  • Stale or invalid
  • Something else…
  • Send Cancel
  bug

  selimfirat/pysad#163 · 1 assignee · Opened Sep 28, 2026

  Maintainers usually reply within 1 day

• Bug in RSHash: far out-of-range instances saturate in the int32 cast in _cell_keys and share every cell key

  Possibly taken @selimfirat claimed this 1 day ago. Open

  ⋯Report a problem with this issue

  • Is something wrong?

  • Already taken
  • Not beginner-friendly
  • Stale or invalid
  • Something else…
  • Send Cancel
  bug

  selimfirat/pysad#162 · 1 assignee · Opened Sep 28, 2026

  Maintainers usually reply within 1 day

• Bug in RSHash: the count-min sketch is an exact dict, so `num_hash_fns` has no effect

  Possibly taken @selimfirat claimed this 2 days ago. Open

  ⋯Report a problem with this issue

  • Is something wrong?

  • Already taken
  • Not beginner-friendly
  • Stale or invalid
  • Something else…
  • Send Cancel
  bug

  selimfirat/pysad#122 · 1 assignee · Opened Sep 27, 2026

  Maintainers usually reply within 1 day

• Inqmad scores ignore later training and are inverted

  Possibly taken @selimfirat claimed this 2 days ago. Open

  ⋯Report a problem with this issue

  • Is something wrong?

  • Already taken
  • Not beginner-friendly
  • Stale or invalid
  • Something else…
  • Send Cancel
  bug

  selimfirat/pysad#120 · 1 assignee · Opened Sep 27, 2026

  Maintainers usually reply within 1 day