def test_bigwig_tiles():
filename = op.join(
'data',
'wgEncodeCaltechRnaSeqHuvecR1x75dTh1014IlnaPlusSignalRep2.bigWig'
)
mean_tile = hgbi.tiles(filename, ['x.0.0'])
mean_mean_tile = hgbi.tiles(filename, ['x.0.0.mean'])
min_tile = hgbi.tiles(filename, ['x.0.0.min'])
max_tile = hgbi.tiles(filename, ['x.0.0.max'])
std_tile = hgbi.tiles(filename, ['x.0.0.std'])
assert mean_tile[0][1]['max_value'] == mean_mean_tile[0][1]['max_value']
assert mean_tile[0][1]['max_value'] > min_tile[0][1]['max_value']
assert max_tile[0][1]['max_value'] > mean_tile[0][1]['max_value']
assert max_tile[0][1]['max_value'] > mean_tile[0][1]['max_value'] + \
std_tile[0][1]['max_value']
min_max_tile = hgbi.tiles(filename, ['x.0.0.minMax'])
whisker_tile = hgbi.tiles(filename, ['x.0.0.whisker'])
mean_val = np.frombuffer(
base64.b64decode(mean_tile[0][1]['dense']),
dtype=mean_tile[0][1]['dtype']
)
min_val = np.frombuffer(
base64.b64decode(min_tile[0][1]['dense']),
dtype=min_tile[0][1]['dtype']
)
max_val = np.frombuffer(
base64.b64decode(max_tile[0][1]['dense']),
dtype=max_tile[0][1]['dtype']
)
std_val = np.frombuffer(
base64.b64decode(std_tile[0][1]['dense']),
dtype=std_tile[0][1]['dtype']
)
min_max_val = np.frombuffer(
base64.b64decode(min_max_tile[0][1]['dense']),
dtype=min_max_tile[0][1]['dtype']
)
whisker_val = np.frombuffer(
base64.b64decode(whisker_tile[0][1]['dense']),
dtype=whisker_tile[0][1]['dtype']
)
assert min_max_val.shape[0] == 2 * mean_val.shape[0]
assert arr_eq_nan(min_max_val[::2], min_val)
assert arr_eq_nan(min_max_val[1::2], max_val)
assert whisker_val.shape[0] == 4 * mean_val.shape[0]
assert arr_eq_nan(whisker_val[::4], min_val)
assert arr_eq_nan(whisker_val[1::4], max_val)
assert arr_eq_nan(whisker_val[2::4], mean_val)
assert arr_eq_nan(whisker_val[3::4], std_val)
def tiles(cls, tileset_dict, tids):
filetype = tileset_dict.get('filetype', "")
if filetype == "bam":
data = bam_tiles.tiles(
tileset_dict.datafile, tids,
index_filename=tileset_dict.index_filename
)
elif filetype == "cooler":
data = cooler_tiles.tiles(tileset_dict['datafile'], tids)
elif filetype == "bigwig":
data = bigwig_tiles.tiles(tileset_dict['datafile'], tids)
else:
data = {'error': f"Unknown tileset filetype: {filetype}"}
return data
def generate_tiles(tileset_tile_ids):
'''
Generate a tiles for the give tile_ids.
All of the tile_ids must come from the same tileset. This function
will determine the appropriate handler this tile given the tileset's
filetype and datatype
Parameters
----------
tileset: tilesets.models.Tileset object
The tileset that the tile ids should be retrieved from
tile_ids: [str,...]
A list of tile_ids (e.g. xyx.0.0.1) identifying the tiles
to be retrieved
Returns
-------
tile_list: [(tile_id, tile_data),...]
A list of tile_id, tile_data tuples
'''
tileset, tile_ids, raw = tileset_tile_ids
if tileset.filetype == 'hitile':
return generate_hitile_tiles(tileset, tile_ids)
elif tileset.filetype == 'beddb':
return hgbe.tiles(tileset.datafile.path, tile_ids)
elif tileset.filetype == 'bed2ddb' or tileset.filetype == '2dannodb':
return generate_bed2ddb_tiles(tileset, tile_ids)
elif tileset.filetype == 'geodb':
return generate_bed2ddb_tiles(tileset, tile_ids, hggo.get_tiles)
elif tileset.filetype == 'hibed':
return generate_hibed_tiles(tileset, tile_ids)
elif tileset.filetype == 'cooler':
return hgco.generate_tiles(tileset.datafile.path, tile_ids)
elif tileset.filetype == 'bigwig':
chromsizes = get_chromsizes(tileset)
return hgbi.tiles(tileset.datafile.path, tile_ids, chromsizes=chromsizes)
elif tileset.filetype == 'multivec':
return generate_1d_tiles(
tileset.datafile.path,
tile_ids,
ctmu.get_single_tile)
elif tileset.filetype == 'imtiles':
return hgim.get_tiles(tileset.datafile.path, tile_ids, raw)
else:
return [(ti, {'error': 'Unknown tileset filetype: {}'.format(tileset.filetype)}) for ti in tile_ids]
def test_bigwig_tiles():
filename = op.join(
'data',
'wgEncodeCaltechRnaSeqHuvecR1x75dTh1014IlnaPlusSignalRep2.bigWig')
hgbi.tiles(filename, ['x.0.0'])
def test_bigwig_tiles():
filename = op.join(
"data",
"wgEncodeCaltechRnaSeqHuvecR1x75dTh1014IlnaPlusSignalRep2.bigWig")
mean_tile = hgbi.tiles(filename, ["x.0.0"])
mean_mean_tile = hgbi.tiles(filename, ["x.0.0.mean"])
min_tile = hgbi.tiles(filename, ["x.0.0.min"])
max_tile = hgbi.tiles(filename, ["x.0.0.max"])
std_tile = hgbi.tiles(filename, ["x.0.0.std"])
assert mean_tile[0][1]["max_value"] == mean_mean_tile[0][1]["max_value"]
assert mean_tile[0][1]["max_value"] > min_tile[0][1]["max_value"]
assert max_tile[0][1]["max_value"] > mean_tile[0][1]["max_value"]
assert (max_tile[0][1]["max_value"] >
mean_tile[0][1]["max_value"] + std_tile[0][1]["max_value"])
min_max_tile = hgbi.tiles(filename, ["x.0.0.minMax"])
whisker_tile = hgbi.tiles(filename, ["x.0.0.whisker"])
mean_val = np.frombuffer(base64.b64decode(mean_tile[0][1]["dense"]),
dtype=mean_tile[0][1]["dtype"])
min_val = np.frombuffer(base64.b64decode(min_tile[0][1]["dense"]),
dtype=min_tile[0][1]["dtype"])
max_val = np.frombuffer(base64.b64decode(max_tile[0][1]["dense"]),
dtype=max_tile[0][1]["dtype"])
std_val = np.frombuffer(base64.b64decode(std_tile[0][1]["dense"]),
dtype=std_tile[0][1]["dtype"])
min_max_val = np.frombuffer(base64.b64decode(min_max_tile[0][1]["dense"]),
dtype=min_max_tile[0][1]["dtype"])
mean_tile = hgbi.tiles(filename, ["x.0.0"])
mean_mean_tile = hgbi.tiles(filename, ["x.0.0.mean"])
min_tile = hgbi.tiles(filename, ["x.0.0.min"])
max_tile = hgbi.tiles(filename, ["x.0.0.max"])
std_tile = hgbi.tiles(filename, ["x.0.0.std"])
assert mean_tile[0][1]["max_value"] == mean_mean_tile[0][1]["max_value"]
assert mean_tile[0][1]["max_value"] > min_tile[0][1]["max_value"]
assert max_tile[0][1]["max_value"] > mean_tile[0][1]["max_value"]
assert (max_tile[0][1]["max_value"] >
mean_tile[0][1]["max_value"] + std_tile[0][1]["max_value"])
min_max_tile = hgbi.tiles(filename, ["x.0.0.minMax"])
whisker_tile = hgbi.tiles(filename, ["x.0.0.whisker"])
mean_val = np.frombuffer(base64.b64decode(mean_tile[0][1]["dense"]),
dtype=mean_tile[0][1]["dtype"])
min_val = np.frombuffer(base64.b64decode(min_tile[0][1]["dense"]),
dtype=min_tile[0][1]["dtype"])
max_val = np.frombuffer(base64.b64decode(max_tile[0][1]["dense"]),
dtype=max_tile[0][1]["dtype"])
std_val = np.frombuffer(base64.b64decode(std_tile[0][1]["dense"]),
dtype=std_tile[0][1]["dtype"])
min_max_val = np.frombuffer(base64.b64decode(min_max_tile[0][1]["dense"]),
dtype=min_max_tile[0][1]["dtype"])
whisker_val = np.frombuffer(base64.b64decode(whisker_tile[0][1]["dense"]),
dtype=whisker_tile[0][1]["dtype"])
assert min_max_val.shape[0] == 2 * mean_val.shape[0]
assert np.allclose(min_max_val[::2], min_val, equal_nan=True)
assert np.allclose(min_max_val[1::2], max_val, equal_nan=True)
assert whisker_val.shape[0] == 4 * mean_val.shape[0]
assert np.allclose(whisker_val[::4], min_val, equal_nan=True)
assert np.allclose(whisker_val[1::4], max_val, equal_nan=True)
assert np.allclose(whisker_val[2::4], mean_val, equal_nan=True)
assert np.allclose(whisker_val[3::4], std_val, equal_nan=True)
def generate_tiles(tileset_tile_ids):
'''
Generate a tiles for the give tile_ids.
All of the tile_ids must come from the same tileset. This function
will determine the appropriate handler this tile given the tileset's
filetype and datatype
Parameters
----------
tileset_tile_ids: tuple
A four-tuple containing the following parameters.
tileset: tilesets.models.Tileset object
The tileset that the tile ids should be retrieved from
tile_ids: [str,...]
A list of tile_ids (e.g. xyx.0.0.1) identifying the tiles
to be retrieved
raw: str or False
The value of the GET request parameter `raw`.
tileset_options: dict or None
An optional dict containing tileset options, including aggregation options.
Returns
-------
tile_list: [(tile_id, tile_data),...]
A list of tile_id, tile_data tuples
'''
tileset, tile_ids, raw, tileset_options = tileset_tile_ids
if tileset.filetype == 'hitile':
return generate_hitile_tiles(tileset, tile_ids)
elif tileset.filetype == 'beddb':
return hgbe.tiles(tileset.datafile.path, tile_ids)
elif tileset.filetype == 'bed2ddb' or tileset.filetype == '2dannodb':
return generate_bed2ddb_tiles(tileset, tile_ids)
elif tileset.filetype == 'geodb':
return generate_bed2ddb_tiles(tileset, tile_ids, hggo.get_tiles)
elif tileset.filetype == 'hibed':
return generate_hibed_tiles(tileset, tile_ids)
elif tileset.filetype == 'cooler':
return hgco.generate_tiles(tileset.datafile.path, tile_ids)
elif tileset.filetype == 'bigwig':
chromsizes = get_chromsizes(tileset)
return hgbi.tiles(tileset.datafile.path,
tile_ids,
chromsizes=chromsizes)
elif tileset.filetype == 'bigbed':
chromsizes = get_chromsizes(tileset)
return hgbb.tiles(tileset.datafile.path,
tile_ids,
chromsizes=chromsizes)
elif tileset.filetype == 'multivec':
return generate_1d_tiles(tileset.datafile.path, tile_ids,
ctmu.get_single_tile, tileset_options)
elif tileset.filetype == 'imtiles':
return hgim.get_tiles(tileset.datafile.path, tile_ids, raw)
elif tileset.filetype == 'bam':
return ctb.tiles(tileset.datafile.path,
tile_ids,
index_filename=tileset.indexfile.path,
max_tile_width=hss.MAX_BAM_TILE_WIDTH)
else:
filetype = tileset.filetype
filepath = tileset.datafile.path
if filetype in hgti.by_filetype:
return hgti.by_filetype[filetype](filepath).tiles(tile_ids)
return [(ti, {
'error':
'Unknown tileset filetype: {}'.format(tileset.filetype)
}) for ti in tile_ids]
def bigwig(filepath, chromsizes=None, uuid=None):
return Tileset(
tileset_info=lambda: hgbi.tileset_info(filepath, chromsizes),
tiles=lambda tids: hgbi.tiles(filepath, tids, chromsizes=chromsizes),
uuid=uuid,
)