def __str__(self):
grouped = Flow(self.components).GroupBy(isa(type=Attribute)).Unboxed()
others = grouped.get(False)
attributes = grouped.get(True)
if not others:
others = ()
if not attributes:
attributes = ()
if not is_unque(map(lambda a: a.name, attributes)):
render_attributes = defaultdict(list)
for each in attributes:
attribute = render_attributes[each.name]
attribute.append(each.components)
attributes = tuple(
Attribute(name, *sum(configs, ()))
for name, configs in render_attributes.items())
return Format.Tag.format(
name=self.name,
indent=Format.Indent * self.indent,
components='\n'.join(map(str, others))
if others else '\n' if self.indent <= 1 else '',
attributes=f' {" ".join(map(str, attributes))}'
if attributes else '')
def test_other():
def block_lambda(e):
e = e + 1
if e < 10:
return e
else:
raise StopIteration
res = Flow(MGenerator(block_lambda, 0)).Take(100).ToList()
assert res.__str__() == res.__repr__()
def DataIOStream(raw_src: Flow):
return (raw_src.Filter(
lambda x: x.endswith('.jpg')) # select jpg files/选取jpg格式文件
.Map(lambda x: os.path.join(train_dir, x)) # 拿到ground truth数据
.Map(data.imread).Map(lambda im: [
im, mixed_noise(im),
gaussian_noise(im),
poisson_noise(im)
] | infix / Map @ img_as_float).Map(to_batch))
def delete(strategy: Strategy):
if strategy.suffixes is all:
log(strategy.dir,
operation='remove',
then_call=partial(shutil.rmtree, ignore_errors=True))
return
(Flow(recur_listdir(strategy.dir)).Filter(endswith(
strategy.suffixes)).Each(
log.within(operation='remove', then_call=os.remove)))
def DataIOStream(raw_src: Flow, num: int):
return (raw_src.Take(num).Filter(
lambda x: x.endswith('.jpg')) # select jpg files/选取jpg格式文件
.Map(lambda x: [os.path.join(train_dir, x)] + [
os.path.join(test_dir, x[:-4] + "_" + str(i) + '.jpg')
for i in range(1, 3)
]) # 将噪声数据和真实数据进行合并
.Map(lambda img_file_names: list(
map(
and_then(
data.imread, # 读取图像
img_as_float), # 浮点数张量 [0, 255]->[0, 1]
img_file_names))).Map(to_batch))
def gen_functions(files):
generated = []
for codes in files:
sources = codes.split('\n')
Flow(sources) \
.Enum() \
.Map(lambda i, x: (i + 1, get_class_value(x)) if x.startswith('@extension_') else None) \
.Filter(lambda x: x) \
.Map(lambda i, value: parser(_parser(token(sources[i]), meta=MetaInfo(), partial=True)
, value)) \
.Filter(lambda x: x) \
.Then(generated.extend)
return '\n'.join(generated)
def ToTupleGenerator(self: dict):
return Flow(((k, v) for k, v in self.items())).ToTuple().Unboxed()
def test_example5():
"""
Example 5:
"""
@extension_class(dict)
def ToTupleGenerator(self: dict):
return Flow(((k, v) for k, v in self.items())).ToTuple().Unboxed()
try:
seq.Take(10).ToTupleGenerator()
except Exception as e:
print(e.args)
"""
NameError: No extension method named `ToTupleGenerator` for builtins.object.
"""
print(seq.Take(10).Zip(seq.Take(10)).ToDict().ToTupleGenerator())
test_example5()
@my_test
def test_extension_byclsname():
@extension_class_name('generator')
def MyNext(self):
return next(self)
test_extension_byclsname()
Flow((i for i in range(10))).MyNext()
def test_Scan():
Flow([2, 3, 5]).Scan(lambda last, now: last + now, 0).ToList()
train_data_size = 500
test_data_size = 100
epochs = 115
batch_group_num = 3
lr = 0.01
loss_fn = torch.nn.MSELoss(size_average=False)
def to_batch(image):
target, *samples = image
return (
np.stack(samples), # X
np.stack([target] * len(samples)))
raw_sources = Flow(os.listdir(train_dir))
def DataIOStream(raw_src: Flow, num: int):
return (raw_src.Take(num).Filter(
lambda x: x.endswith('.jpg')) # select jpg files/选取jpg格式文件
.Map(lambda x: [os.path.join(train_dir, x)] + [
os.path.join(test_dir, x[:-4] + "_" + str(i) + '.jpg')
for i in range(1, 3)
]) # 将噪声数据和真实数据进行合并
.Map(lambda img_file_names: list(
map(
and_then(
data.imread, # 读取图像
img_as_float), # 浮点数张量 [0, 255]->[0, 1]
img_file_names))).Map(to_batch))
}]
(Flow(
map(Strategy.new, [
*additional, {
"./GraphEngine/ffi": suffixes1 + suffixes2 + suffixes3 + ['.sig']
}, {
"./GraphEngine.egg-info": all
}, {
"./__pycache__": all
}, {
"./build": all
}, {
"./dist": all
}, {
'storage': all
}, {
"GraphEngine/ffi/storage": all
}, {
'cache': all
}, {
'GraphEngine/ffi/A': all
}, {
'GraphEngine/ffi/B': all
}, {
'GraphEngine/ffi/composite-helper': all
}, {
'GraphEngine/ffi/write_ahead_log': all
}
])).Each(delete))
from matplotlib import pyplot as plt
from math import exp
from itertools import cycle
import dill
model = torch.load('model', pickle_module=dill)
print('load model')
# 数据下载地址见: http://smartdsp.xmu.edu.cn/cvpr2017.html
# download data-sets here: http://smartdsp.xmu.edu.cn/cvpr2017.html
train_dir = './rainy_image_dataset/ground truth'
test_dir = './rainy_image_dataset/rainy image'
train_data_size = 500
test_data_size = 100
lr = 0.1
raw_sources = Flow(os.listdir(train_dir))
def DataIOStream(raw_src: Flow, num: int):
return (raw_src.Take(num).Then(cycle).Filter(
lambda x: x.endswith('.jpg')) # select jpg files/选取jpg格式文件
.Map(lambda x: [os.path.join(train_dir, x)] + [
os.path.join(test_dir, x[:-4] + "_" + str(i) + '.jpg')
for i in range(1, 3)
]) # 将噪声数据和真实数据进行合并
.Map(lambda img_file_names: list(
map(
and_then(
data.imread, # 读取图像
img_as_float), # 浮点数张量 [0, 255]->[0, 1]
img_file_names))).Map(to_batch))
import GraphEngine.ffi
import Trinity
import os, clr
from linq import Flow
from Trinity.FFI import Agent
import Trinity
IncludeDirectory = os.path.join(GraphEngine.__path__[0], 'ffi')
StorageRoot = os.path.abspath('storage')
TSLCodeGenExeLocation = os.path.join(GraphEngine.__path__[0], 'Command',
'Trinity.TSL.CodeGen.exe')
DotNetExeLocation = 'dotnet.exe'
Agent.Configure(IncludeDirectory, StorageRoot, TSLCodeGenExeLocation,
DotNetExeLocation, 10, 10, 10)
Trinity.Global.LocalStorage.LoadStorage()
Agent.Initialize()
if not os.path.exists(r"storage\composite-helper\Trinity.Extension.abc.dll"):
tsl_path = os.path.abspath('./tests/tsl')
Agent.LoadTSL(*[tsl_path, tsl_path, "abc", None])
Flow(
Trinity.Global.StorageSchema.CellDescriptors).Each(lambda cell_desc: print(
f'{cell_desc.TypeName}{list(cell_desc.GetFieldNames())}'))
Trinity.Global.LocalStorage.SaveStorage()
Agent.Uninitialize()
def test_ArgSorted():
Flow([(1, 2), (2, 3), (3, 2)]).ArgSorted(lambda x, y: x + y)
Flow([3, 2, 1]).ArgSorted()
Flow([(1, 1), (2, 2), (3, 1)]).ArgSorted(by=lambda a, b: a * b).ToList()
def test_Sorted():
Flow([(1, 2), (2, 3), (3, 2)]).Sorted(lambda x, y: x + y)
Flow([1, 2, 3]).Sorted().Sorted(by=lambda x: -x)
def test_Zip():
Flow([(1, 2), (2, 3), (3, 2)]).Zip([(1, 2), (2, 2), (3, 3)])
def test_Aggregate():
Flow([1, 2, 3, 4, 5]).Aggregate(max, min, sum).ToTuple()
def test_Each():
Flow([(1, 2), (2, 3), (3, 2)]).Each(lambda x, y: x + y)
def test_Filter():
Flow([(1, 2), (2, 3), (3, 2)]).Filter(lambda x, y: x + y)
Flow([(1, 1), (2, 2), (3, 2)]).Filter(lambda x, y: x is y).Filter().Filter(
lambda x: x != (3, 2)).All()
def call():
global seq
seq = Flow(MGenerator(lambda x: x + 1, start_elem=0)) # [0..\infty]
func.__globals__['seq'] = seq
func()
def test_Union():
Flow({(1, 1), (2, 2), (3, 3)}).Union([(1, 2), (2, 2), (3, 3)])
def test_Group():
Flow([(1, 2), (2, 3), (3, 2)]).Group(lambda x, y: x + y).ToTuple()
Flow([1, 1, 2, 3, 3]).Group().Map(lambda _: (len(_), len(_))).Group(
lambda a, b: a * b).ToTuple()
def test_Then():
Flow([(1, 2), (2, 3)]).Then(lambda x: x)
Flow([(1, 2), (2, 3)]).Then(lambda x, y: x + y)
def test_GroupBy():
Flow([(1, 2), (2, 3), (3, 2)]).GroupBy(lambda x, y: x + y).ToTuple()
Flow([1, 1, 1]).GroupBy().ToList()
def test_Reversed():
Flow([(1, 2), (2, 2), (3, 3)]).Reversed()
def getitems(self):
return Flow(
self.result.nodes).Map(lambda n: self.generateitem(n)).Unboxed()
def test_Intersects():
Flow({(1, 1), (2, 2), (3, 3)}).Intersects([(1, 2), (2, 2), (3, 3)])
def getitems(self):
return Flow(self.data["features"]).Map(
lambda n: self.generateitem(n)).Unboxed()
from .core.ParserC import *
from .__release_info__ import *
from Redy.Tools.PathLib import Path
import os
from linq import Flow
RBNF_HOME = 'RBNF_HOME'
_root_dir = Path(__file__).parent()
home = Path(os.environ.get(RBNF_HOME, '~/.rbnf'))
if RBNF_HOME not in os.environ:
os.environ[RBNF_HOME] = home.__str__()
if not home.exists():
Flow(_root_dir.into('rbnf_libs').list_dir()).each(
lambda it: it.move_to(home))
def test_Sum():
Flow([(1, 2), (2, 3), (3, 2)]).Sum(lambda x, y: x + y)
