Can not get same results from tutorial desc_2.1.1_paul.ipynb

[ichobits]

I can not get the same results from the tutorial desc_2.1.1_paul.ipynb and desc_2.0.3_paul.ipynb.
https://eleozzr.github.io/desc/tutorial.html

my desc_2.1.1 env:
conda create the python==3.6.13

Name	Version	Build
absl-py	0.12.0	pypi_0
anndata	0.7.5	pypi_0
anyio	2.2.0	py36ha15d459_0
argon2-cffi	20.1.0	py36h68aa20f_2
astunparse	1.6.3	pypi_0
async_generator	1.1	py_0
attrs	20.3.0	pyhd3deb0d_0
babel	2.9.0	pyhd3deb0d_0
backports	1	py_2
backports.functools_lru_cache	1.6.3	pyhd8ed1ab_0
bleach	3.3.0	pyh44b312d_0
cached-property	1.5.2	pypi_0
cachetools	4.2.1	pypi_0
certifi	2020.12.5	py36ha15d459_1
cffi	1.14.5	py36he58ceb7_0
chardet	4.0.0	pypi_0
click	7.1.2	pypi_0
colorama	0.4.4	pyh9f0ad1d_0
contextvars	2.4	py_0
cycler	0.10.0	pypi_0
dataclasses	0.8	pyh787bdff_0
decorator	4.4.2	pypi_0
defusedxml	0.7.1	pyhd8ed1ab_0
desc	2.1.1	pypi_0
entrypoints	0.3	pyhd8ed1ab_1003
flatbuffers	1.12	pypi_0
gast	0.3.3	pypi_0
get-version	2.1	pypi_0
google-auth	1.28.0	pypi_0
google-auth-oauthlib	0.4.4	pypi_0
google-pasta	0.2.0	pypi_0
grpcio	1.32.0	pypi_0
h5py	2.10.0	pypi_0
idna	2.1	pypi_0
immutables	0.15	py36h68aa20f_0
importlib-metadata	3.10.0	py36ha15d459_0
ipykernel	5.5.3	py36hfacbf0b_0
ipython	5.8.0	py36_1
ipython_genutils	0.2.0	py_1
jinja2	2.11.3	pyh44b312d_0
joblib	1.0.1	pypi_0
json5	0.9.5	pyh9f0ad1d_0
jsonschema	3.2.0	pyhd8ed1ab_3
jupyter-packaging	0.7.12	pyhd8ed1ab_0
jupyter_client	6.1.12	pyhd8ed1ab_0
jupyter_core	4.7.1	py36ha15d459_0
jupyter_server	1.5.1	py36ha15d459_0
jupyterlab	3.0.12	pyhd8ed1ab_0
jupyterlab_pygments	0.1.2	pyh9f0ad1d_0
jupyterlab_server	2.4.0	pyhd8ed1ab_0
keras-preprocessing	1.1.2	pypi_0
kiwisolver	1.3.1	pypi_0
legacy-api-wrap	1.2	pypi_0
libsodium	1.0.18	h8d14728_1
llvmlite	0.36.0	pypi_0
m2w64-gcc-libgfortran	5.3.0	6
m2w64-gcc-libs	5.3.0	7
m2w64-gcc-libs-core	5.3.0	7
m2w64-gmp	6.1.0	2
m2w64-libwinpthread-git	5.0.0.4634.697f757	2
markdown	3.3.4	pypi_0
markupsafe	1.1.1	py36h68aa20f_3
matplotlib	3.3.4	pypi_0
mistune	0.8.4	py36h68aa20f_1003
msys2-conda-epoch	20160418	1
natsort	7.1.1	pypi_0
nbclassic	0.2.6	pyhd8ed1ab_0
nbclient	0.5.3	pyhd8ed1ab_0
nbconvert	6.0.7	py36ha15d459_3
nbformat	5.1.3	pyhd8ed1ab_0
nest-asyncio	1.5.1	pyhd8ed1ab_0
networkx	2.5.1	pypi_0
notebook	6.3.0	py36ha15d459_0
numba	0.53.1	pypi_0
numexpr	2.7.3	pypi_0
numpy	1.19.5	pypi_0
oauthlib	3.1.0	pypi_0
opt-einsum	3.3.0	pypi_0
packaging	20.9	pyh44b312d_0
pandas	1.1.5	pypi_0
pandoc	2.13	h8ffe710_0
pandocfilters	1.4.2	py_1
patsy	0.5.1	pypi_0
pickleshare	0.7.5	py_1003
pillow	8.2.0	pypi_0
pip	21.0.1	pyhd8ed1ab_0
prometheus_client	0.10.0	pyhd8ed1ab_0
prompt_toolkit	1.0.15	py_1
protobuf	3.15.7	pypi_0
pyasn1	0.4.8	pypi_0
pyasn1-modules	0.2.8	pypi_0
pycparser	2.2	pyh9f0ad1d_2
pydot	1.4.2	pypi_0
pygments	2.8.1	pyhd8ed1ab_0
pynndescent	0.5.2	pypi_0
pyparsing	2.4.7	pyh9f0ad1d_0
pyrsistent	0.17.3	py36h68aa20f_2
python	3.6.13	h39d44d4_0_cpython
python-dateutil	2.8.1	py_0
python_abi	3.6	1_cp36m
pytz	2021.1	pyhd8ed1ab_0
pywin32	300	py36h68aa20f_0
pywinpty	0.5.7	py36h9f0ad1d_1
pyzmq	22.0.3	py36h1d5d788_1
requests	2.25.1	pypi_0
requests-oauthlib	1.3.0	pypi_0
rsa	4.7.2	pypi_0
scanpy	1.5.1	pypi_0
scikit-learn	0.24.1	pypi_0
scipy	1.5.4	pypi_0
seaborn	0.11.1	pypi_0
send2trash	1.5.0	py_0
setuptools	54.2.0	pypi_0
setuptools-scm	6.0.1	pypi_0
simplegeneric	0.8.1	py_1
six	1.15.0	pyh9f0ad1d_0
sniffio	1.2.0	py36ha15d459_1
statsmodels	0.12.2	pypi_0
stdlib-list	0.8.0	pypi_0
tables	3.6.1	pypi_0
tensorboard	2.4.1	pypi_0
tensorboard-plugin-wit	1.8.0	pypi_0
tensorflow	2.4.1	pypi_0
tensorflow-estimator	2.4.0	pypi_0
termcolor	1.1.0	pypi_0
terminado	0.9.4	py36ha15d459_0
testpath	0.4.4	py_0
texttable	1.6.3	pypi_0
threadpoolctl	2.1.0	pypi_0
tornado	6.1	py36h68aa20f_1
tqdm	4.60.0	pypi_0
traitlets	4.3.3	py36h9f0ad1d_1
typing_extensions	3.7.4.3	py_0
umap-learn	0.5.1	pypi_0
urllib3	1.26.4	pypi_0
vc	14.2	hb210afc_4
vs2015_runtime	14.28.29325	h5e1d092_4
wcwidth	0.2.5	pyh9f0ad1d_2
webencodings	0.5.1	py_1
werkzeug	1.0.1	pypi_0
wheel	0.36.2	pyhd3deb0d_0
wincertstore	0.2	py36ha15d459_1006
winpty	0.4.3	4
wrapt	1.12.1	pypi_0
zeromq	4.3.4	h0e60522_0
zipp	3.4.1	pyhd8ed1ab_0

---

Then, I follow the tutorial desc_2.1.1_paul.ipynb.

1.different cell matrix

desc_2.1.1_paul.ipynb.
AnnData object with n_obs × n_vars = 2730 × 999
obs: 'paul15_clusters', 'celltype', 'celltype2', 'desc_0.8', 'desc_1.0'
var: 'highly_variable', 'means', 'dispersions', 'dispersions_norm', 'mean', 'std'
uns: 'iroot', 'log1p', 'prob_matrix0.8', 'prob_matrix1.0'
obsm: 'X_Embeded_z0.8', 'X_tsne', 'X_tsne0.8', 'X_Embeded_z1.0', 'X_tsne1.0'

my script:
AnnData object with n_obs × n_vars = 2730 × 1000
obs: 'paul15_clusters', 'celltype', 'celltype2', 'desc_0.8', 'desc_1.0'
var: 'highly_variable', 'means', 'dispersions', 'dispersions_norm', 'mean', 'std'
uns: 'iroot', 'log1p', 'prob_matrix0.8', 'prob_matrix1.0'
obsm: 'X_Embeded_z0.8', 'X_tsne', 'X_tsne0.8', 'X_Embeded_z1.0', 'X_tsne1.0'

2.different t-sen
picture
desc_2.1.1_paul.ipynb.
https://ibb.co/GsPv38j

my script:
https://ibb.co/NxhHgq0

What is the cause of these differences?

[eleozzr]

Maybe due to the version of scanpy. Sometimes scanpy cannot exactly identify 1000 hvgs even you set n_hvg=1000.