def test_inv():
df = tibble(x=1, y=2)
out = df >> select(~f.x)
assert out.columns.tolist() == ["y"]
df = tibble(x=True)
out = df >> mutate(y=~f.x)
assert out.y.tolist() == [False]
def gene_name_conversion(
genes,
species,
infmt,
outfmt,
notfound,
):
"""Convert gene names using MyGeneInfo
Args:
genes: A sequence of genes
species: The species to limit the query
Supported: human, mouse, rat, fruitfly, nematode, zebrafish,
thale-cress, frog and pig
infmt: What's the original gene name format
Available fields
https://docs.mygene.info/en/latest/doc/query_service.html#available-fields
outfmt: What's the target gene name format
notfound: What to do if a conversion cannot be done.
use-query: Ignore the conversion and use the original name
skip: Ignore the conversion and skip the entire row in input file
error: Report error
Returns:
A dataframe with two columns, query and `outfmt`.
"""
out = (mygene.querymany(
genes,
scopes=infmt,
fields=outfmt,
as_dataframe=True,
df_index=False,
species=species,
) >> group_by(f.query) >> arrange(desc(f._score)) >> slice_head(1) >>
select(~c(f._id, f._score, f.notfound)))
if isinstance(outfmt, str):
outfmt = [of.strip() for of in outfmt.split(",")]
out = tibble(query=genes) >> left_join(out, by=f.query)
if notfound == "use-query":
out = out >> mutate(
across(
outfmt,
lambda col, query: if_else(is_na(col), query, col),
query=f.query,
))
elif notfound == "error" and any(is_na(out[outfmt[0]])):
nagenes = out >> filter(is_na(f[outfmt[0]])) >> pull(f.query)
raise QueryGenesNotFound(nagenes)
elif notfound == "skip":
out = out >> filter(~is_na(f[outfmt[0]]))
return out
class TruvariBench(TruvariBench):
requires = DownloadList
input_data = lambda ch: tibble(
compvcf=expand_dir(ch, pattern="input*.vcf.gz"),
basevcf=expand_dir(ch, pattern="multi*.vcf.gz").iloc[0, 0],
)
envs = {
"ref":
str(
Path(__file__).parent.parent.parent / "data" / "reference" /
"hg19" / "chrs.fa"),
}
def test_and_or():
df = tibble(x=1, y=2, z=3, w=4)
out = df >> select(c(f.x, f.y) & c(f.y, f.z))
assert out.columns.tolist() == ["y"]
out = df >> mutate(a=f.x & f.y)
assert out.a.tolist() == [True]
out = df >> mutate(a=True & f.y)
assert out.a.tolist() == [True]
out = df >> select(c(f.x, f.y) | c(f.y, f.z))
assert out.columns.tolist() == ["x", "y", "z"]
def test_glimpse_html_df():
df = tibble(x=f[:20], y=[str(i) for i in range(20)])
g = glimpse(df, 100)
out = g._repr_html_()
assert "<table>" in out
def test_glimpse_str_gf():
df = tibble(x=f[:10], y=[str(i) for i in range(10)]) >> group_by(f.y)
out = repr(glimpse(df))
assert "Groups: y [10]" in out
def test_glimpse_str_nest_df():
df = tibble(x=f[:10], y=f[10:20]) >> nest(data=~f.x)
out = str(glimpse(df))
assert "Rows: 10" in out
assert "Columns: 2" in out
assert "<DF 1x1>, <DF 1x1>" in out
def test_glimpse_str_df():
df = tibble(x=f[:10], y=[str(i) for i in range(10)])
out = str(glimpse(df))
assert "Rows: 10" in out
assert "Columns: 2" in out
assert "0, 1, 2" in out
from datar.all import tibble, flatten
from pipen import Proc
from biopipen.ns.web import Download
from biopipen.ns.plot import Heatmap
from biopipen.core.testing import get_pipeline
Heatmap = Proc.from_proc(
Heatmap,
requires=Download,
input_data=lambda ch: tibble(infile=ch, annofiles=[flatten(ch)]),
envs={
"globals": "data = head(data, 100)",
"args": {
"right_annotation":
"""r:rowAnnotation(
Boxplot = anno_boxplot(as.matrix(head(annos, 100)), outline = F)
)"""
}
})
def pipeline():
return get_pipeline(__file__).set_starts(Download).set_data([
"https://www.ncbi.nlm.nih.gov/geo/download/"
"?acc=GSE179367"
"&format=file"
"&file=GSE179367%5Fgene%5Fcount%2Ereal%2Etxt%2Egz",
])
def testing(pipen):
def test_or_():
df = tibble(x=1, y=2, z=3)
out = df >> select(c(f.x, f.y) | [f.y, f.z])
assert_frame_equal(out, tibble(x=1, y=2, z=3))
def test_right_recycle_to_left():
df = tibble(x=[True, False])
out = mutate(df, y=f.x | True)
assert_frame_equal(out, tibble(x=[True, False], y=[True, True]))
def test_rowwise_gets_rowwise():
df = tibble(x=[1, 2, 3], y=[4, 5, 6]).rowwise()
out = mutate(df, z=1 + f.y, w=-f.x, t=+f.y)
assert out.z.is_rowwise
assert out.w.is_rowwise
def test_op_getattr():
df = tibble(x=[1, 2], y=[1, -3])
out = mutate(df, z=(f.x * f.y).abs())
assert_frame_equal(out, tibble(x=[1, 2], y=[1, -3], z=[1, 6]))
def test_undefined_op():
df = tibble(x=[1, 2], y=[1, 3])
out = mutate(df, z=f.x * f.y)
assert_frame_equal(out, tibble(x=[1, 2], y=[1, 3], z=[1, 6]))
def test_ne():
df = tibble(x=[1, 2], y=[1, 3])
out = mutate(df, z=f.x != f.y)
assert_frame_equal(out, tibble(x=[1, 2], y=[1, 3], z=[False, True]))
out = mutate(df, z=f.x.size != f.y.size)
assert_frame_equal(out, tibble(x=[1, 2], y=[1, 3], z=[False, False]))
def test_apply_dtypes():
df = tibble(x=[1.0, 2.0])
apply_dtypes(df, True)
assert is_integer(df.x)
def test_neg():
df = tibble(x=1, y=2)
out = df >> select(-f[:1])
assert out.columns.tolist() == ["y"]
from pipen import Proc
from biopipen.ns.web import Download
from biopipen.ns.misc import Str2File
from biopipen.ns.gsea import FGSEA
from datar.all import flatten, tibble, select
from biopipen.core.testing import get_pipeline
FGSEA = Proc.from_proc(
FGSEA,
requires=[Download, Str2File],
input_data=lambda ch1, ch2: tibble(
*flatten(ch1),
ch2,
_name_repair="minimal",
) >> select(0, 2, 1),
envs={"clscol": "Group", "classes": ["MMP9", "CTRL"]}
)
def pipeline():
return get_pipeline(__file__).set_starts(Download, Str2File).set_data(
[
"https://www.ncbi.nlm.nih.gov/geo/download/"
"?acc=GSE179367"
"&format=file"
"&file=GSE179367%5Fgene%5Fcount%2Ereal%2Etxt%2Egz",
"https://www.genepattern.org/tutorial/linkedFiles/"
"export_gnf.GENE_SYMBOL.gmt",
],
[(
def test_neg():
df = tibble(x=[1, 2])
out = df >> mutate(y=-f.x)
assert_frame_equal(out, tibble(x=[1, 2], y=[-1, -2]))
