def runProcedure():
rez = mldb.perform("PUT", "/v1/procedures/rocket_science", [], conf)
mldb.log(json.loads(rez["response"]))
assert rez["statusCode"] == 201
rez = mldb.perform("GET", "/v1/query", [["q", "select * from toy_tsne"]])
return json.loads(rez["response"])
def test_deadloack(self):
# MLDB-621
try:
mldb.perform(
"PUT", "/v1/procedures/q", [], {
"type": "serial",
"params": {
"steps": [{
"id": "q",
"type": "null"
}]
}
})
except Exception:
pass
def create_dataset_and_query(type, date, quantum = None):
if quantum:
ds1 = mldb.create_dataset({
'type': type,
'id': 'beh',
'params' : {
'timeQuantumSeconds' : quantum
}
})
else:
ds1 = mldb.create_dataset({
'type': type,
'id': 'beh'
})
sometime = datetime.strptime(date, '%Y-%m-%dT%H:%M:%S.%fZ')
ds1.record_row('row', [['column', 1, sometime]])
ds1.commit()
response = mldb.perform('GET', '/v1/query', [['q', 'select * from beh']])['response']
mldb.perform('DELETE', '/v1/datasets/beh', [])
return json.loads(response)[0]['columns'][0][2]
# This file is part of MLDB. Copyright 2015 mldb.ai inc. All rights reserved.
import json
from mldb import mldb
result = mldb.perform(
"PUT", "/v1/plugins/mirror", [], {
"type": "python",
"params": {
"source": {
"routes":
"""
from mldb import mldb
mldb.log(str(request.rest_params))
mldb.log(str(request.payload))
request.set_return({
"args": request.rest_params,
"payload": request.payload
})
"""
}
}
})
assert result["statusCode"] < 400, result["response"]
successes = 0
rtn = mldb.perform("POST", "/v1/plugins/mirror/routes/pwet")
if rtn["response"] == '{"args":[],"payload":"null\\n"}':
successes += 1
def perform(*args, **kwargs):
res = mldb.perform(*args, **kwargs)
assert res['statusCode'] in [200, 201], str(res)
return res
def query(sql):
result = mldb.perform("GET", "/v1/query",
[["q", sql], ["format", "table"]], {})
assert result["statusCode"] < 400, result["response"]
return result["response"]
import json
from mldb import mldb
csv_conf = {
"type": "import.text",
"params": {
'dataFileUrl':
"https://raw.githubusercontent.com/datacratic/mldb-pytanic-plugin/master/titanic_train.csv",
"outputDataset": {
"id": "titanic_raw",
},
"runOnCreation": True,
}
}
mldb.perform("PUT", "/v1/procedures/csv_proc", [], csv_conf)
result = mldb.perform(
"PUT", "/v1/procedures/titanic_train_scorer", [], {
"type": "classifier.experiment",
"params": {
"experimentName": "titanic",
"keepArtifacts": True,
"inputData": """
select
{Sex, Age, Fare, Embarked, Parch, SibSp, Pclass} as features,
label
from titanic_raw
""",
"modelFileUrlPattern": "file://tmp/MLDB-1304.cls",
"algorithm": "bbdt",
# This file is part of MLDB. Copyright 2015 mldb.ai inc. All rights reserved.
import json
from mldb import mldb
result = mldb.perform(
"PUT", "/v1/datasets/rcp", [], {
'type': 'beh',
'params': {
'dataFileUrl': 'https://s3.amazonaws.com/public-mldb-ai/rcp.beh'
}
})
assert result["statusCode"] < 400, result["response"]
mldb.log(result)
out = mldb.perform("GET", "/v1/query",
[["q", "select * from rcp order by rowName() limit 20"],
["format", "sparse"]], {})
mldb.log(out)
result = mldb.perform(
"PUT", "/v1/procedures/svd1", [], {
'type': 'svd.train',
'params': {
'trainingData': {
'from': {
'id': 'rcp'
},
'limit': 1000
"id": "iris_dataset"
},
"select":
'"petal width", "petal length", "sepal length", "sepal width"'
},
"columnOutputDataset": {
"type": "sparse.mutable",
"id": "svd_iris_col"
},
"rowOutputDataset": {
"id": "iris_svd_row",
'type': "embedding"
},
"numSingularValues": 4,
"numDenseBasisVectors": 2
}
}
r = mldb.perform("PUT", svd_procedure, [], svd_config)
print(r)
print((r["statusCode"]))
r = mldb.perform("PUT", svd_procedure + "/runs/1", [], {})
if not 300 > r["statusCode"] >= 200:
print(r)
request.set_return("FAILURE")
print((r["statusCode"]))
else:
request.set_return("success")
# This file is part of MLDB. Copyright 2015 mldb.ai inc. All rights reserved.
import json
from mldb import mldb
# Dataset with one column and one row
ds1 = mldb.create_dataset({'type': 'sparse.mutable', 'id': 'ds1'})
for i in range(1):
ds1.record_row('row_' + str(i), [['x', i, 1438895158]])
ds1.commit()
# Simple query of "x, rowName() as rowname" passed through our function
res = mldb.perform('GET', '/v1/query',
[['q', 'SELECT x, 1 as y, rowName() as rowname FROM ds1']])
response = json.loads(res['response'])
expected = response
mldb.log(response)
assert res['statusCode'] == 200
# Same query but in a transform procedure
res = mldb.perform(
'PUT', '/v1/procedures/poil', [], {
'type': 'transform',
'params': {
'inputData': {
'select': 'x, 1 as y, rowName() as rowname',
'from': {
'id': 'ds1'
dataset_config["id"] = "dataset3"
dataset = mldb.create_dataset(dataset_config)
dataset.record_row("row1", [["col1", "a", now], ["otherRow", "row1", now]])
dataset.record_row("row2", [["col2", "b", now], ["otherRow", "row2", now]])
dataset.commit()
#MLDB-1454 rownames in WHERE regular
mldb.log("MLDB-1454 rownames in WHERE regular")
expected = [["_rowName", "col1"], ["row1", "a"]]
res = mldb.perform("GET", "/v1/query", [[
"q", """
SELECT *
FROM dataset1
WHERE dataset1.col1 = 'a'
"""
], ['format', 'table']])
assert res["statusCode"] == 200
assert json.loads(res['response']) == expected
##
expected = [["_rowName", "col1"], ["row1", "a"]]
res = mldb.perform("GET", "/v1/query", [[
"q", """
SELECT *
FROM dataset1 as blah
WHERE blah.col1 = 'a'
"""
#
# beh-binary-mutable-save-test.py
# Mich, 2015-12-14
# This file is part of MLDB. Copyright 2015 mldb.ai inc. All rights reserved.
#
import tempfile
import os
import json
from mldb import mldb
res = mldb.perform('PUT', '/v1/datasets/myDataset', [], {
'type' : 'beh.binary.mutable'
})
assert res['statusCode'] == 201, str(res)
res = mldb.perform('POST', '/v1/datasets/myDataset/rows', [], {
'rowName' : 'row1',
'columns' : [
['colA', 1, 0]
]
})
assert res['statusCode'] == 200, str(res)
res = mldb.perform('POST', '/v1/datasets/myDataset/rows', [], {
'rowName' : 'row2',
'columns' : [
['colB', 1, 0],
]
})
assert res['statusCode'] == 200, str(res)
###
import json
from mldb import mldb
conf = {
"source":
"""
from mldb import mldb
mldb.log("patate")
mldb.log({"patate":2.44})
mldb.log(["patate", "pwel"])
mldb.log(25)
mldb.log("a", "b", 2)
mldb.log()
"""
}
rtn = mldb.perform("POST", "/v1/types/plugins/python/routes/run", [], conf)
jsRtn = json.loads(rtn["response"])
mldb.log(jsRtn)
assert jsRtn["logs"][0]["c"] == "patate"
assert jsRtn["logs"][1]["c"] == "{\n \"patate\": 2.44\n}"
assert jsRtn["logs"][2]["c"] == "[\n \"patate\",\n \"pwel\"\n]"
assert jsRtn["logs"][3]["c"] == "25"
assert jsRtn["logs"][4]["c"] == "a b 2"
#assert jsRtn["logs"][5]["c"] == ""
request.set_return("success")
ds2.record_row('row_0', [['a', 0, 0]])
ds2.record_row('row_1', [['a', 3, 0]])
ds2.record_row('row_2', [['a', 6, 0]])
ds2.record_row('row_3', [['a', 9, 0]])
ds2.record_row('row_4', [['a', 12, 0]])
ds1.commit()
ds2.commit()
# This function returns the row corresponding to the `id` passed from the first
# dataset
res = mldb.perform(
'PUT', '/v1/functions/patate', [], {
'type': 'sql.query',
'params': {
'query': 'select * from ds1 where rowName() = $id'
}
})
check_res(res, 201)
# Make sure the function works properly
res = mldb.perform('GET', '/v1/functions/patate/application',
[['input', '{"id":"row_2"}']], {})
response = check_res(res, 200)
assert response["output"]['x'] == 2
assert response["output"]['y'] == 4
# Check the function's info output
test_dataset = mldb.create_dataset({
"type": "sparse.mutable",
"id": "test_dataset"
})
for i in range(50):
feat1 = random.randint(1, 20)
feat2 = random.randint(1, 100)
test_dataset.record_row("u%d" % i,
[["feat1", feat1, now], ["feat2", feat2, now]])
test_dataset.commit()
mldb.log(
json.loads(
mldb.perform("GET", "/v1/query",
[["q", "SELECT * from dataset limit 10"]])["response"]))
config = {
"type": "classifier.train",
"params": {
"trainingData": """
SELECT {
feat1, feat2
} as features,
label
FROM dataset""",
"algorithm": "glz",
"configuration": {
"glz": {
"type": "glz",
"verbosity": 3,
def applyFunction(row):
rez = mldb.perform("GET", "/v1/functions/tsne_embed/application", [['input', {'embedding':row}]])
mldb.log(rez)
return rez
# This file is part of MLDB. Copyright 2016 mldb.ai inc. All rights reserved.
#####
# Testing if the error responses are all formatted in the same way
# Guy Dumais, January 27, 2016
# Copyright mldb.ai inc 2016
####
import json, random, datetime
from mldb import mldb
rez = mldb.perform("GET", "/v1/query", [['q', 'sele']], {})
mldb.log(rez)
response = json.loads(rez['response'])
mldb.log(response)
assert 'error' in response, 'expected an error message in the response'
assert 'httpCode' in response, 'expected an http code in the response'
rez = mldb.perform("GET", "/v1/query", [['q', 'select h()']], {})
mldb.log(rez)
response = json.loads(rez['response'])
mldb.log(response)
assert 'error' in response, 'expected an error message in the response'
assert 'httpCode' in response, 'expected an http code in the response'
rez = mldb.perform("GET", "/v1/querry", [], {})
mldb.log(rez)
response = json.loads(rez['response'])
mldb.log(response)
assert 'error' in response, 'expected an error message in the response'
assert 'httpCode' in response, 'expected an http code in the response'
