Sam/css refresh (#161) (#634)

initialize list outside of not args.suppress_plots block. (#159)

initialize list outside of not args.suppress_plots block.


move amino acid plot into not args.suppress_plots block.

Remove nesting of if block for plotting amino acid plots

Fix type in suppress_plots




Add tool name for report generation in CRISPRessoPooledWGSCompare

Fix issue 612, coding_seq filename too long (#152)

Add discovery of Python packages (#151)

Fix Docker entrypoint

Attempt to fix Circle CI pip install

Run each CRISPResso command in the conda environment

Update --base_editor_output parameter name for Circle CI

Update tests Makefile and batch expected output

Fix QWC inference across amplicons (#137)

Mckay/be plot improvements (#136)

trying to get the figure to fit nicely, increased
element size to 100

custom figsize to display without cutting off


increased figsize in report template

Allow messages to be served in CLI reports (#134) (#583)

Fix deletion at second position (#131)

Fix bug when there is a deletion starting at the second position


This bug only happens when a deletion starts are the second position, before the
fix, it would report that the deletion started at the first position. It is
fixed now, so deletions at the second position are reported correctly.

Update CRISPRessoCOREResources.c due to change in .pyx

Add tests for find_indels_substitutions for deletions at the end

Fix 1bp deletions at the end, and off by one error


This ensures that when a deletion occurs at the end of a read, the entire
deletion is accounted for.

Update CRISPRessoCOREResources.c to reflect fixes for deletions at the end of alignments

Add extra asserts to deletion checks

Point to new test branch

Reafctor deletion_coordinates to go past the end of the string for deletions at
the end of the sequence

Point tests to master

Allow messages to be served in CLI reports

Point to cole/messages test branch

Point tests back to master

point to tests branch

typo

testing github actions

remove test

point tests to master




Update inferred QWC tests to reflect correct intended behavior

Fix inferring QWC to match intended behavior

Add more test cases and fix bug discovered in single bp QWC

Add even more test cases testing indels outside the QWC

Point tests to cole/fix-qwc-deletion

update plotly.js (#138)

Change order of amplicon inference alignment so that 1st amplicon is the reference


This makes a difference because it changes the values of s1inds, and therefore
the value of the inferred quantification window coordinates.

Point integration tests back to master

Update CHANGELOG.md




Update setup.py and pyproject.toml to find all modules and packages

Convert pyproject.toml to Unix file endings




added func: get_short_seq_id, that truncates long sequence, and adds unique identifier.

point to tests branch

mmm

Sam/pixi integration tests (#156)

Guardrail dropdowns (#155)

Add discovery of Python packages (#151)

Fix Docker entrypoint

Attempt to fix Circle CI pip install

Run each CRISPResso command in the conda environment

Update --base_editor_output parameter name for Circle CI

Update tests Makefile and batch expected output

Fix QWC inference across amplicons (#137)

Mckay/be plot improvements (#136)

trying to get the figure to fit nicely, increased
element size to 100

custom figsize to display without cutting off


increased figsize in report template

Allow messages to be served in CLI reports (#134) (#583)

Fix deletion at second position (#131)

Fix bug when there is a deletion starting at the second position


This bug only happens when a deletion starts are the second position, before the
fix, it would report that the deletion started at the first position. It is
fixed now, so deletions at the second position are reported correctly.

Update CRISPRessoCOREResources.c due to change in .pyx

Add tests for find_indels_substitutions for deletions at the end

Fix 1bp deletions at the end, and off by one error


This ensures that when a deletion occurs at the end of a read, the entire
deletion is accounted for.

Update CRISPRessoCOREResources.c to reflect fixes for deletions at the end of alignments

Add extra asserts to deletion checks

Point to new test branch

Reafctor deletion_coordinates to go past the end of the string for deletions at
the end of the sequence

Point tests to master

Allow messages to be served in CLI reports

Point to cole/messages test branch

Point tests back to master

point to tests branch

typo

testing github actions

remove test

point tests to master




Update inferred QWC tests to reflect correct intended behavior

Fix inferring QWC to match intended behavior

Add more test cases and fix bug discovered in single bp QWC

Add even more test cases testing indels outside the QWC

Point tests to cole/fix-qwc-deletion

update plotly.js (#138)

Change order of amplicon inference alignment so that 1st amplicon is the reference


This makes a difference because it changes the values of s1inds, and therefore
the value of the inferred quantification window coordinates.

Point integration tests back to master

Update CHANGELOG.md




Update setup.py and pyproject.toml to find all modules and packages

Convert pyproject.toml to Unix file endings




Guardrail dropdowns

Revert changes to .c files

Closed guardrail warning by default

.c fix

Revert .c files to match master



Revert .c files back to master

Update CHANGELOG.md

Update CHANGELOG.md




Run integration tests

Fix pixi env

Remove pro from pixi.toml

Pull branches

Adding Pro token

Add pip install kaleido for pro tests

Pro compare

Pin plotly

Pin kaleido

Fix kaleido pin

Fix pixi plotly pin

Ruff fix

Update integration_tests.yml


Remove branch ref for Pro

Run the diffs and print the results in integration tests

Point tests to master




lint

Mckay/upsetplot port (#154)

new arg

Diff plots (#158)

Run diffs on the plots in integration tests

Correct diff-plots now

Pin matplotlib in test dependencies and update running integration tests

Fix cross-pollution of integration tests

Install MS core fonts in CI for consistent plot rendering


CRISPResso2 requests Arial as the primary font via matplotlib’s
font.sans-serif setting. Without MS core fonts installed, the CI
runner falls back to a different font, causing upset plots (and
potentially others) to exceed the RMSE image comparison threshold.
This matches the Dockerfile which already installs
ttf-mscorefonts-installer. Also clears matplotlib’s font cache
so it discovers the newly installed fonts.

Remove matplotlib pin in tests and don’t run diff-plots with Pro

Reorder base edit plots and update caption in Compare

Rebuild pixi fontconfig cache after MS font installation in CI


The pixi/conda environment has its own fontconfig with a separate
font cache (.pixi/envs/test/var/cache/fontconfig/). When the pixi
environment is restored from actions/cache, this fontconfig cache
is stale and doesn’t know about the MS core fonts installed via
apt-get. Running the system fc-cache only updates the system
fontconfig, not the pixi one.
Fix: run ‘pixi run -e test – fc-cache -f’ to rebuild the pixi
fontconfig cache so that matplotlib’s findSystemFonts() discovers
Arial via the pixi fc-list.

Update base editing caption in compare

Sort Compare allele plots by aligned sequence as well

Attempt to fix sorting of compare allele plots

Sort the df being passed into the plot in compare!

Sort the bottom compare plots too!

Make the tiebreaking logic in compare enrichment consistent

Cache MS fonts in pixi

Point integration tests back to master and revert pixi.toml

Fix mutating args directly in compare

Make the parameter names consistent across the code

Point tests to master




Add CRISPRessoPro advertisement banner to top and bottom

Replace cup with svg

Bug Fix: cut_point not in exon_positions (#150)

VCF Output (#128)

vcf file writing initial concept

splitting vcf into testable functions, adding tests

adding unit tests, adding nullcontext to vcf writing file

fixed first insertion bug

Mckay/base edit plot (#119)

Cole/plot fixes (#121)

fix setting of 99%ile in negative direction (deletions)

need another break

When reading CRISPRessoPooled amplicon file, only skip lines if they start with the comment character(#)

Implement new overwrite_crispresso_options in CRISPRessoPooled to only add non-default commands when propagating

In CRISPRessoPooled implement multiplexing for sub-runs and create a run_name that is filename-safe and separate from a display name.

Add arg for display_name

Change display name name to Display Name

CRISPRessoWGS fix unpickleable partial error

Move complete message to after crispresso cup to allow for json parsing of status.json

Update CRISPRessoWGS to allow for run names

Messages to users about counting reads in input (sometimes this takes a while for large samples)

Cast guardrail values as ints because division casts them as floats

Remove invalid escape of _ when writing to JSON status file

Fix documentatio for CRISPRessoShared.check_if_failed and remove extraneous whitespace

Point to updated test branch

Fix Cython bug

Read version from toml file

Fix import error

update version

Update integration_tests.yml


Remove Native Merge test temporarily

Pin fastp version

Run native merge test and point tests to master




Additional plot fixes (#122)

fix setting of 99%ile in negative direction (deletions)

need another break

When reading CRISPRessoPooled amplicon file, only skip lines if they start with the comment character(#)

Implement new overwrite_crispresso_options in CRISPRessoPooled to only add non-default commands when propagating

In CRISPRessoPooled implement multiplexing for sub-runs and create a run_name that is filename-safe and separate from a display name.

Add arg for display_name

Change display name name to Display Name

CRISPRessoWGS fix unpickleable partial error

Move complete message to after crispresso cup to allow for json parsing of status.json

Update CRISPRessoWGS to allow for run names

Messages to users about counting reads in input (sometimes this takes a while for large samples)

Cast guardrail values as ints because division casts them as floats

Remove invalid escape of _ when writing to JSON status file

Fix documentatio for CRISPRessoShared.check_if_failed and remove extraneous whitespace

Point to updated test branch

Fix Cython bug

Read version from toml file

Move printing header info to after console log level setting in main function

Add import for importlib.metadata to CRISPRessoShared.py

Standardize intermediate file names for CRISPRessoPooled info and fastq files

CRISPRessoPooled avoid gzipping nonexistant files

Remove warning if no config file

Allow none for custom_colors in CRISPRessoPlot

read version from toml file

Make fig_filename_root default to None, in which case the figure is shown interactively - e.g. in a jupyter notebook

Print tool description after logging level has been set

update testRelease.sh script

Don’t rerun if the ‘verbosity’ value has changed.

CRISPResso core will not rerun if there are changes in the ‘debug’, ‘n_processes’, and ‘verbosity’ arguments

Verbosity levels <=1 are set to 1, >= 4 are set to 4.

update for future pd.read_json deprecation

Fix import error

update version

Update integration_tests.yml


Remove Native Merge test temporarily

Pin fastp version

Run native merge test and point tests to master

Point tests to cole/plot_fixes branch

Point tests back to master





added Kendall’s code




added args
added fig 10i to report
added tests for data prep


added upsetplot to Dockerfile

added save_png to plot args

removed unused args

remove unused variable

cole’s comments

removed unused function

comments

comments

updated test_be_df with UNMODIFIED row

remove comment

removed extra args
fixed name

Update help string for --base_editor_consider_changes_outside_qw

Point tests back to master




further vcf tweaks/work

vcf dynamic alt map and vcf line generation with unit testing

updating tests

Add test case for second element deletion

Remove unnecessary imports and add whitspace

Add mini integration tests for alt_map and vcf_line

Update unit tests to account for bug in find_indels_substitutions

Add failing test case for insertion and deletion occurring at the same position


The alt map incorrectly separate the deletion and insertion to different
positions.

Cast dict_keys to be a list

Fix test case to reflect correct position of deletion

Add checks to the VCF file

Add upsetplot as a dependency in setup.py (#130)

Fix position of deletion in test case

Fix off by one error for deletions, update tests and fix deletion at start


It turns out that when there is a deletion at the start of the sequence, the
correct way to handle it is to provide the last base after the deletion. Source:
https://bioinformatics.stackexchange.com/questions/2476/how-to-represent-a-deletion-at-position-1-in-a-vcf-file

Update tests to convince myself that insertion and deletion handling is working

Remove duplicate writing of allele frequency table

Refactor get_allele_row and make vcf_output not dependent on write_detailed_allele_table

Extract out construction of df_alleles in unit tests

Write test to illustrate bug when you have an insertion then a deletion

Refactor unit tests to use create_df_alleles

Fix test_build_alt_map_mixed test

Refactor create_df_alleles to support number of reads


This fixes test_build_alt_map_substitutions

Fix test_build_alt_map_insertions test

Allow amplicon name to be passed to create_df_alleles

Fix test_upsert_edit_del_and_ins

Fix deletion at start of amplicon

Fix bug when there is a deletion starting at the second position


This bug only happens when a deletion starts are the second position, before the
fix, it would report that the deletion started at the first position. It is
fixed now, so deletions at the second position are reported correctly.

Fix test to reflect full deletion

Add tests for find_indels_substitutions for deletions at the end

Fix 1bp deletions at the end, and off by one error


This ensures that when a deletion occurs at the end of a read, the entire
deletion is accounted for.

Fix for representing deletions at the end of a sequence

Fix bug where deletions that extend to the end of a sequence fail

Remove qwc_indexes

Properly implement handling of deletions that start at the beginning of a sequence

Properly account for multiple delete_start events

Update test to reflect proper ref_positions and other attributes

Fix QWC inference across amplicons (#137)

Mckay/be plot improvements (#136)

trying to get the figure to fit nicely, increased
element size to 100

custom figsize to display without cutting off


increased figsize in report template

Allow messages to be served in CLI reports (#134) (#583)

Fix deletion at second position (#131)

Fix bug when there is a deletion starting at the second position


This bug only happens when a deletion starts are the second position, before the
fix, it would report that the deletion started at the first position. It is
fixed now, so deletions at the second position are reported correctly.

Update CRISPRessoCOREResources.c due to change in .pyx

Add tests for find_indels_substitutions for deletions at the end

Fix 1bp deletions at the end, and off by one error


This ensures that when a deletion occurs at the end of a read, the entire
deletion is accounted for.

Update CRISPRessoCOREResources.c to reflect fixes for deletions at the end of alignments

Add extra asserts to deletion checks

Point to new test branch

Reafctor deletion_coordinates to go past the end of the string for deletions at
the end of the sequence

Point tests to master

Allow messages to be served in CLI reports

Point to cole/messages test branch

Point tests back to master

point to tests branch

typo

testing github actions

remove test

point tests to master




Update inferred QWC tests to reflect correct intended behavior

Fix inferring QWC to match intended behavior

Add more test cases and fix bug discovered in single bp QWC

Add even more test cases testing indels outside the QWC

Point tests to cole/fix-qwc-deletion

update plotly.js (#138)

Change order of amplicon inference alignment so that 1st amplicon is the reference


This makes a difference because it changes the values of s1inds, and therefore
the value of the inferred quantification window coordinates.

Point integration tests back to master

Update CHANGELOG.md




Fix a bug for --bam_output when there are unaligned reads (#144)

Fix BAM output when there are unaligned reads

Point tests to cole/unaligned-reads

Update CHANGELOG.md

Fix typo in tests branch name

Point tests back to master

Add VCF testing and verification design document


Documents the strategy for integration testing the –vcf_output feature
using golden file comparison with syn-gen synthetic data.

Change name of output VCF file

Point tests to vcf-parameters

Add writers module refactor design document


Documents plan to move VCF output code from CRISPRessoUtilities.py
to a new writers/ module structure.

Refactor VCF code into writers/ module

Move CRISPRessoUtilities.py to writers/vcf.py to establish a pattern
for organizing output writers. Update imports in CRISPRessoCORE.py
and relocate tests to test_writers/test_vcf/.

Remove the FORMAT column from VCF and add the contig length to VCF headers

Initial VCF simplication writing each edit on its own line

Further simplification of VCF output code

Left align deletions for VCF output

Left-normalize insertions for VCF output

Update CLAUDE.md with design_docs

Update Github Actions with Pooled Prime Editing and VCF basic and VCF Prime Edit Basic

Update integration tests to run in parallel

Add VCF path to crispresso2_info

Point the tests back to cole/vcf-parameters

Always save the conda env cache in Github Actions

Debug: add set +e and per-target error reporting in integration tests


setup-miniconda@v3 injects ‘set -eo pipefail’ into ~/.profile, which
is sourced by bash -l. This causes make commands to fail silently.
Adding set +e at script start and explicit per-target error annotations.

Attempt to fix conda cache in GitHub Actions

Update CHANGELOG.md and parameter descriptions

Remove unused nullcontext import

Remove extra blank lines after amplicon_coordinates validation

Document that –vcf_output implies –write_detailed_allele_table

Use max amplicon length for VCF contig headers and warn on collision


When multiple amplicons map to the same chromosome, the contig header
now uses the max amplicon length instead of silently overwriting with
the last one seen. A warning is logged when this occurs.

Fix VCF POS off-by-one for start-of-amplicon deletions when pos > 1

For deletions starting at position 0 of the amplicon, the VCF POS was
computed as pos-1 instead of pos. The max(1,…) clamp accidentally
masked this when pos=1 (all existing tests). With pos>1, VCF POS
pointed one base before the amplicon.
Fix: when start==0, set left_index=pos directly instead of
pos+start-1.

Fix multi-insertion bug: use alignment index to extract inserted bases

_edits_from_insertions treated the second element of insertion_coordinates
as an alignment index, but it is actually a reference position. These
coincide for single-insertion reads but diverge when prior insertions
shift alignment indices.
Fix: convert right_anchor_ref_pos to an alignment index via
ref_positions.index(), then extract the size characters before it.
Also:

Rename misleading variables (right_anchor_ref_pos was actually left
anchor, aligned_start was actually right anchor ref pos)
Add comment on defensive ref_len branch explaining it is currently
unreachable
Update LEFT_NORMALIZATION.md to correctly describe the
insertion_coordinates format


Add clarifying example to help text of –amplicon_coordinates

Remove planning documents

Point tests back to master

Update PR number in CHANGELOG.md

Cache the conda environment for pytest as well




clamp cut point within exon region
if not in list, set to 0

point tests to new branch

fix cut_point calculation

Cole’s changes

Add discovery of Python packages (#151)

Fix Docker entrypoint

Attempt to fix Circle CI pip install

Run each CRISPResso command in the conda environment

Update --base_editor_output parameter name for Circle CI

Update tests Makefile and batch expected output

Fix QWC inference across amplicons (#137)

Mckay/be plot improvements (#136)

trying to get the figure to fit nicely, increased
element size to 100

custom figsize to display without cutting off


increased figsize in report template

Allow messages to be served in CLI reports (#134) (#583)

Fix deletion at second position (#131)

Fix bug when there is a deletion starting at the second position


This bug only happens when a deletion starts are the second position, before the
fix, it would report that the deletion started at the first position. It is
fixed now, so deletions at the second position are reported correctly.

Update CRISPRessoCOREResources.c due to change in .pyx

Add tests for find_indels_substitutions for deletions at the end

Fix 1bp deletions at the end, and off by one error


This ensures that when a deletion occurs at the end of a read, the entire
deletion is accounted for.

Update CRISPRessoCOREResources.c to reflect fixes for deletions at the end of alignments

Add extra asserts to deletion checks

Point to new test branch

Reafctor deletion_coordinates to go past the end of the string for deletions at
the end of the sequence

Point tests to master

Allow messages to be served in CLI reports

Point to cole/messages test branch

Point tests back to master

point to tests branch

typo

testing github actions

remove test

point tests to master




Update inferred QWC tests to reflect correct intended behavior

Fix inferring QWC to match intended behavior

Add more test cases and fix bug discovered in single bp QWC

Add even more test cases testing indels outside the QWC

Point tests to cole/fix-qwc-deletion

update plotly.js (#138)

Change order of amplicon inference alignment so that 1st amplicon is the reference


This makes a difference because it changes the values of s1inds, and therefore
the value of the inferred quantification window coordinates.

Point integration tests back to master

Update CHANGELOG.md




Update setup.py and pyproject.toml to find all modules and packages

Convert pyproject.toml to Unix file endings




Update CHANGELOG.md

Fix ruff linting

Diff plots (#158)

Run diffs on the plots in integration tests

Correct diff-plots now

Pin matplotlib in test dependencies and update running integration tests

Fix cross-pollution of integration tests

Install MS core fonts in CI for consistent plot rendering


CRISPResso2 requests Arial as the primary font via matplotlib’s
font.sans-serif setting. Without MS core fonts installed, the CI
runner falls back to a different font, causing upset plots (and
potentially others) to exceed the RMSE image comparison threshold.
This matches the Dockerfile which already installs
ttf-mscorefonts-installer. Also clears matplotlib’s font cache
so it discovers the newly installed fonts.

Remove matplotlib pin in tests and don’t run diff-plots with Pro

Reorder base edit plots and update caption in Compare

Rebuild pixi fontconfig cache after MS font installation in CI


The pixi/conda environment has its own fontconfig with a separate
font cache (.pixi/envs/test/var/cache/fontconfig/). When the pixi
environment is restored from actions/cache, this fontconfig cache
is stale and doesn’t know about the MS core fonts installed via
apt-get. Running the system fc-cache only updates the system
fontconfig, not the pixi one.
Fix: run ‘pixi run -e test – fc-cache -f’ to rebuild the pixi
fontconfig cache so that matplotlib’s findSystemFonts() discovers
Arial via the pixi fc-list.

Update base editing caption in compare

Sort Compare allele plots by aligned sequence as well

Attempt to fix sorting of compare allele plots

Sort the df being passed into the plot in compare!

Sort the bottom compare plots too!

Make the tiebreaking logic in compare enrichment consistent

Cache MS fonts in pixi

Point integration tests back to master and revert pixi.toml

Fix mutating args directly in compare

Fix issue 612, coding_seq filename too long (#152)

Add discovery of Python packages (#151)

Fix Docker entrypoint

Attempt to fix Circle CI pip install

Run each CRISPResso command in the conda environment

Update --base_editor_output parameter name for Circle CI

Update tests Makefile and batch expected output

Fix QWC inference across amplicons (#137)

Mckay/be plot improvements (#136)

trying to get the figure to fit nicely, increased
element size to 100

custom figsize to display without cutting off


increased figsize in report template

Allow messages to be served in CLI reports (#134) (#583)

Fix deletion at second position (#131)

Fix bug when there is a deletion starting at the second position


This bug only happens when a deletion starts are the second position, before the
fix, it would report that the deletion started at the first position. It is
fixed now, so deletions at the second position are reported correctly.

Update CRISPRessoCOREResources.c due to change in .pyx

Add tests for find_indels_substitutions for deletions at the end

Fix 1bp deletions at the end, and off by one error


This ensures that when a deletion occurs at the end of a read, the entire
deletion is accounted for.

Update CRISPRessoCOREResources.c to reflect fixes for deletions at the end of alignments

Add extra asserts to deletion checks

Point to new test branch

Reafctor deletion_coordinates to go past the end of the string for deletions at
the end of the sequence

Point tests to master

Allow messages to be served in CLI reports

Point to cole/messages test branch

Point tests back to master

point to tests branch

typo

testing github actions

remove test

point tests to master




Update inferred QWC tests to reflect correct intended behavior

Fix inferring QWC to match intended behavior

Add more test cases and fix bug discovered in single bp QWC

Add even more test cases testing indels outside the QWC

Point tests to cole/fix-qwc-deletion

update plotly.js (#138)

Change order of amplicon inference alignment so that 1st amplicon is the reference


This makes a difference because it changes the values of s1inds, and therefore
the value of the inferred quantification window coordinates.

Point integration tests back to master

Update CHANGELOG.md




Update setup.py and pyproject.toml to find all modules and packages

Convert pyproject.toml to Unix file endings




added func: get_short_seq_id, that truncates long sequence, and adds unique identifier.

point to tests branch

mmm

Sam/pixi integration tests (#156)

Guardrail dropdowns (#155)

Add discovery of Python packages (#151)

Fix Docker entrypoint

Attempt to fix Circle CI pip install

Run each CRISPResso command in the conda environment

Update --base_editor_output parameter name for Circle CI

Update tests Makefile and batch expected output

Fix QWC inference across amplicons (#137)

Mckay/be plot improvements (#136)

trying to get the figure to fit nicely, increased
element size to 100

custom figsize to display without cutting off


increased figsize in report template

Allow messages to be served in CLI reports (#134) (#583)

Fix deletion at second position (#131)

Fix bug when there is a deletion starting at the second position


This bug only happens when a deletion starts are the second position, before the
fix, it would report that the deletion started at the first position. It is
fixed now, so deletions at the second position are reported correctly.

Update CRISPRessoCOREResources.c due to change in .pyx

Add tests for find_indels_substitutions for deletions at the end

Fix 1bp deletions at the end, and off by one error


This ensures that when a deletion occurs at the end of a read, the entire
deletion is accounted for.

Update CRISPRessoCOREResources.c to reflect fixes for deletions at the end of alignments

Add extra asserts to deletion checks

Point to new test branch

Reafctor deletion_coordinates to go past the end of the string for deletions at
the end of the sequence

Point tests to master

Allow messages to be served in CLI reports

Point to cole/messages test branch

Point tests back to master

point to tests branch

typo

testing github actions

remove test

point tests to master




Update inferred QWC tests to reflect correct intended behavior

Fix inferring QWC to match intended behavior

Add more test cases and fix bug discovered in single bp QWC

Add even more test cases testing indels outside the QWC

Point tests to cole/fix-qwc-deletion

update plotly.js (#138)

Change order of amplicon inference alignment so that 1st amplicon is the reference


This makes a difference because it changes the values of s1inds, and therefore
the value of the inferred quantification window coordinates.

Point integration tests back to master

Update CHANGELOG.md




Update setup.py and pyproject.toml to find all modules and packages

Convert pyproject.toml to Unix file endings




Guardrail dropdowns

Revert changes to .c files

Closed guardrail warning by default

.c fix

Revert .c files to match master



Revert .c files back to master

Update CHANGELOG.md

Update CHANGELOG.md




Run integration tests

Fix pixi env

Remove pro from pixi.toml

Pull branches

Adding Pro token

Add pip install kaleido for pro tests

Pro compare

Pin plotly

Pin kaleido

Fix kaleido pin

Fix pixi plotly pin

Ruff fix

Update integration_tests.yml


Remove branch ref for Pro

Run the diffs and print the results in integration tests

Point tests to master




lint

Mckay/upsetplot port (#154)

new arg

Diff plots (#158)

Run diffs on the plots in integration tests

Correct diff-plots now

Pin matplotlib in test dependencies and update running integration tests

Fix cross-pollution of integration tests

Install MS core fonts in CI for consistent plot rendering


CRISPResso2 requests Arial as the primary font via matplotlib’s
font.sans-serif setting. Without MS core fonts installed, the CI
runner falls back to a different font, causing upset plots (and
potentially others) to exceed the RMSE image comparison threshold.
This matches the Dockerfile which already installs
ttf-mscorefonts-installer. Also clears matplotlib’s font cache
so it discovers the newly installed fonts.

Remove matplotlib pin in tests and don’t run diff-plots with Pro

Reorder base edit plots and update caption in Compare

Rebuild pixi fontconfig cache after MS font installation in CI


The pixi/conda environment has its own fontconfig with a separate
font cache (.pixi/envs/test/var/cache/fontconfig/). When the pixi
environment is restored from actions/cache, this fontconfig cache
is stale and doesn’t know about the MS core fonts installed via
apt-get. Running the system fc-cache only updates the system
fontconfig, not the pixi one.
Fix: run ‘pixi run -e test – fc-cache -f’ to rebuild the pixi
fontconfig cache so that matplotlib’s findSystemFonts() discovers
Arial via the pixi fc-list.

Update base editing caption in compare

Sort Compare allele plots by aligned sequence as well

Attempt to fix sorting of compare allele plots

Sort the df being passed into the plot in compare!

Sort the bottom compare plots too!

Make the tiebreaking logic in compare enrichment consistent

Cache MS fonts in pixi

Point integration tests back to master and revert pixi.toml

Fix mutating args directly in compare

Make the parameter names consistent across the code

Point tests to master




Remove sgRNA parameters from plot_amino_acid_heatmap

Update CHANGELOG.md

Point tests to master




Styles update

Final reports fixes

cut_idx = None (#162)

cut_idx = None

cut point fix

remove comments

docstring, remove unused var

ruff

space

space

remove error handling. the starting index is guaranteed to be preset in ref_positions.

Point tests to master




Invalidate stale pixi cache (v2) to fix libcblas.so.3 ImportError

The cached .pixi directory became corrupted/stale, causing numpy 2.4.3
to fail with ‘libcblas.so.3: cannot open shared object file’. Bumping
the cache key version forces a fresh environment install.

Center header


Co-authored-by: Samuel Nichols Snic9004@gmail.com
Co-authored-by: mbowcut2 55161542+mbowcut2@users.noreply.github.com
Co-authored-by: Claude Opus 4.6 noreply@anthropic.com
Co-authored-by: Trevor Martin trevormartinj7@gmail.com
Co-authored-by: kclem k.clement.dev@gmail.com