Skip to content

Updates for the adapted nuisance est for the IV-type score (PLR) & the new IV-type score for PLIV #12

New issue

Have a question about this project? Sign up for a free GitHub account to open an issue and contact its maintainers and the community.

Sign up for GitHub

By clicking “Sign up for GitHub”, you agree to our terms of service and privacy statement. We’ll occasionally send you account related emails.

Already on GitHub? Sign in to your account

Merged
merged 3 commits into from
Jun 14, 2022
Merged

Updates for the adapted nuisance est for the IV-type score (PLR) & the new IV-type score for PLIV #12

Show file tree
Hide file tree
Changes from all commits
Commits
Show all changes
3 commits
Select commit Hold shift + click to select a range
6025631
adaptions for the iv type score and renamed nuisance functions
MalteKurz Jun 14, 2022
4772182
Merge branch 'master' of github.com:DoubleML/doubleml-py-vs-r into m-…
MalteKurz Jun 14, 2022
274caf9
adaptions and extensions for iv-type scores
MalteKurz Jun 14, 2022
File filter

Filter by extension

Filter by extension
Conversations
Failed to load comments.
Loading
Jump to
Jump to file
Failed to load files.
Loading
Diff view
Diff view
38 changes: 28 additions & 10 deletions doubleml_py_vs_r/tests/_utils_pyvsr.py
View file
Open in desktop
Original file line number Diff line number Diff line change
Expand Up @@ -27,17 +27,23 @@ def export_smpl_split_to_r(smpls):

f <- function(data, score, dml_procedure, n_rep, smpls_for_r) {
data = data.table(data)
mlmethod_l = 'regr.lm'
mlmethod_m = 'regr.lm'
mlmethod_g = 'regr.lm'
if (score == "IV-type") {
mlmethod_g = 'regr.lm'
} else {
mlmethod_g = NULL
}

Xnames = names(data)[names(data) %in% c("y", "d") == FALSE]
data_ml = double_ml_data_from_data_frame(data, y_col = "y",
d_cols = "d", x_cols = Xnames)

double_mlplr_obj = DoubleMLPLR$new(data_ml,
n_folds = 2,
ml_g = mlmethod_g,
ml_l = mlmethod_l,
ml_m = mlmethod_m,
ml_g = mlmethod_g,
dml_procedure = dml_procedure,
score = score)
smpls = list()
Expand All @@ -63,9 +69,14 @@ def export_smpl_split_to_r(smpls):

f <- function(data, score, dml_procedure, train_ids, test_ids) {
data = data.table(data)
mlmethod_g = 'regr.lm'
mlmethod_l = 'regr.lm'
mlmethod_m = 'regr.lm'
mlmethod_r = 'regr.lm'
if (score == "IV-type") {
mlmethod_g = 'regr.lm'
} else {
mlmethod_g = NULL
}

Xnames = names(data)[names(data) %in% c("y", "d", "Z1") == FALSE]
data_ml = double_ml_data_from_data_frame(data, y_col = "y",
Expand All @@ -74,9 +85,10 @@ def export_smpl_split_to_r(smpls):

double_mlpliv_obj = DoubleMLPLIV$new(data_ml,
n_folds = 2,
ml_g = mlmethod_g,
ml_l = mlmethod_l,
ml_m = mlmethod_m,
ml_r = mlmethod_r,
ml_g = mlmethod_g,
dml_procedure = dml_procedure,
score = score)

Expand All @@ -98,7 +110,7 @@ def export_smpl_split_to_r(smpls):

f <- function(data, score, dml_procedure, train_ids, test_ids) {
data = data.table(data)
mlmethod_g = 'regr.lm'
mlmethod_l = 'regr.lm'
mlmethod_m = 'regr.lm'
mlmethod_r = 'regr.lm'

Expand All @@ -110,7 +122,7 @@ def export_smpl_split_to_r(smpls):

double_mlpliv_obj = DoubleML:::DoubleMLPLIV.partialX(data_ml,
n_folds = 2,
ml_g = mlmethod_g,
ml_l = mlmethod_l,
ml_m = mlmethod_m,
ml_r = mlmethod_r,
dml_procedure = dml_procedure,
Expand Down Expand Up @@ -166,7 +178,7 @@ def export_smpl_split_to_r(smpls):

f <- function(data, score, dml_procedure, train_ids, test_ids) {
data = data.table(data)
mlmethod_g = 'regr.lm'
mlmethod_l = 'regr.lm'
mlmethod_m = 'regr.lm'
mlmethod_r = 'regr.lm'

Expand All @@ -178,7 +190,7 @@ def export_smpl_split_to_r(smpls):

double_mlpliv_obj = DoubleML:::DoubleMLPLIV.partialXZ(data_ml,
n_folds = 2,
ml_g = mlmethod_g,
ml_l = mlmethod_l,
ml_m = mlmethod_m,
ml_r = mlmethod_r,
dml_procedure = dml_procedure,
Expand Down Expand Up @@ -270,9 +282,14 @@ def export_smpl_split_to_r(smpls):
train_ids, test_ids,
cluster_var1, cluster_var2=NULL) {
data = data.table(data)
mlmethod_g = 'regr.lm'
mlmethod_l = 'regr.lm'
mlmethod_m = 'regr.lm'
mlmethod_r = 'regr.lm'
if (score == "IV-type") {
mlmethod_g = 'regr.lm'
} else {
mlmethod_g = NULL
}

if (is.null(cluster_var2)) cluster_vars = cluster_var1 else cluster_vars = c(cluster_var1, cluster_var2)
Xnames = names(data)[names(data) %in% c("Y", "D", "Z", cluster_vars) == FALSE]
Expand All @@ -283,9 +300,10 @@ def export_smpl_split_to_r(smpls):

double_mlpliv_obj = DoubleMLPLIV$new(data_ml,
n_folds = 2,
ml_g = mlmethod_g,
ml_l = mlmethod_l,
ml_m = mlmethod_m,
ml_r = mlmethod_r,
ml_g = mlmethod_g,
dml_procedure = dml_procedure,
score = score)

Expand Down
2 changes: 1 addition & 1 deletion doubleml_py_vs_r/tests/test_iivm_pyvsr.py
View file
Open in desktop
Original file line number Diff line number Diff line change
Expand Up @@ -41,7 +41,7 @@ def dml_iivm_pyvsr_fixture(generate_data_iivm, score, dml_procedure):

dml_iivm_obj = dml.DoubleMLIIVM(obj_dml_data,
ml_g, ml_m, ml_r,
n_folds,
n_folds=n_folds,
dml_procedure=dml_procedure)

np.random.seed(3141)
Expand Down
2 changes: 1 addition & 1 deletion doubleml_py_vs_r/tests/test_irm_pyvsr.py
View file
Open in desktop
Original file line number Diff line number Diff line change
Expand Up @@ -40,7 +40,7 @@ def dml_irm_pyvsr_fixture(generate_data_irm, score, dml_procedure):

dml_irm_obj = dml.DoubleMLIRM(obj_dml_data,
ml_g, ml_m,
n_folds,
n_folds=n_folds,
score=score,
dml_procedure=dml_procedure)

Expand Down
40 changes: 28 additions & 12 deletions doubleml_py_vs_r/tests/test_pliv_multiway_cluster_pyvsr.py
View file
Open in desktop
Original file line number Diff line number Diff line change
Expand Up @@ -13,29 +13,40 @@
pandas2ri.activate()


@pytest.fixture(scope='module',
params=['partialling out', 'IV-type'])
def score(request):
return request.param


@pytest.fixture(scope='module',
params=['dml1', 'dml2'])
def dml_procedure(request):
return request.param


@pytest.fixture(scope='module')
def dml_pliv_twoway_cluster_pyvsr_fixture(generate_data_pliv_two_way_cluster, dml_procedure):
def dml_pliv_twoway_cluster_pyvsr_fixture(generate_data_pliv_two_way_cluster, score, dml_procedure):
n_folds = 2

# collect data
obj_dml_data = generate_data_pliv_two_way_cluster

# Set machine learning methods for g, m & r
# Set machine learning methods for l, m, & r
learner = LinearRegression()
ml_g = clone(learner)
ml_l = clone(learner)
ml_m = clone(learner)
ml_r = clone(learner)
if score == 'IV-type':
ml_g = clone(learner)
else:
ml_g = None

np.random.seed(3141)
dml_pliv_obj = dml.DoubleMLPLIV(obj_dml_data,
ml_g, ml_m, ml_r,
n_folds,
ml_l, ml_m, ml_r, ml_g,
n_folds=n_folds,
score=score,
dml_procedure=dml_procedure)
print(obj_dml_data)
dml_pliv_obj.fit()
Expand All @@ -44,7 +55,7 @@ def dml_pliv_twoway_cluster_pyvsr_fixture(generate_data_pliv_two_way_cluster, dm
all_train, all_test = export_smpl_split_to_r(dml_pliv_obj.smpls[0])

r_dataframe = pandas2ri.py2rpy(obj_dml_data.data)
res_r = r_MLPLIV_multiway_cluster(r_dataframe, 'partialling out', dml_procedure,
res_r = r_MLPLIV_multiway_cluster(r_dataframe, score, dml_procedure,
all_train, all_test,
obj_dml_data.cluster_cols[0],
obj_dml_data.cluster_cols[1])
Expand All @@ -70,30 +81,35 @@ def test_dml_pliv_twoway_cluster_pyvsr_se(dml_pliv_twoway_cluster_pyvsr_fixture)


@pytest.fixture(scope='module')
def dml_pliv_one_cluster_pyvsr_fixture(generate_data_pliv_one_way_cluster, dml_procedure):
def dml_pliv_one_cluster_pyvsr_fixture(generate_data_pliv_one_way_cluster, score, dml_procedure):
n_folds = 2

# collect data
obj_dml_data = generate_data_pliv_one_way_cluster

# Set machine learning methods for g, m & r
# Set machine learning methods for l, m & r
learner = LinearRegression()
ml_g = clone(learner)
ml_l = clone(learner)
ml_m = clone(learner)
ml_r = clone(learner)
if score == 'IV-type':
ml_g = clone(learner)
else:
ml_g = None

np.random.seed(3141)
dml_pliv_obj = dml.DoubleMLPLIV(obj_dml_data,
ml_g, ml_m, ml_r,
n_folds,
ml_l, ml_m, ml_r, ml_g,
n_folds=n_folds,
score=score,
dml_procedure=dml_procedure)
dml_pliv_obj.fit()

# fit the DML model in R
all_train, all_test = export_smpl_split_to_r(dml_pliv_obj.smpls[0])

r_dataframe = pandas2ri.py2rpy(obj_dml_data.data.drop(columns='cluster_var_j'))
res_r = r_MLPLIV_multiway_cluster(r_dataframe, 'partialling out', dml_procedure,
res_r = r_MLPLIV_multiway_cluster(r_dataframe, score, dml_procedure,
all_train, all_test,
obj_dml_data.cluster_cols[0])

Expand Down
43 changes: 24 additions & 19 deletions doubleml_py_vs_r/tests/test_pliv_pyvsr.py
View file
Open in desktop
Original file line number Diff line number Diff line change
Expand Up @@ -15,7 +15,7 @@


@pytest.fixture(scope='module',
params=['partialling out'])
params=['partialling out', 'IV-type'])
def score(request):
return request.param

Expand All @@ -33,16 +33,21 @@ def dml_pliv_pyvsr_fixture(generate_data_pliv, score, dml_procedure):
# collect data
obj_dml_data = generate_data_pliv

# Set machine learning methods for g, m & r
# Set machine learning methods for l, m, r & g
learner = LinearRegression()
ml_g = clone(learner)
ml_l = clone(learner)
ml_m = clone(learner)
ml_r = clone(learner)
if score == 'IV-type':
ml_g = clone(learner)
else:
ml_g = None

np.random.seed(3141)
dml_pliv_obj = dml.DoubleMLPLIV(obj_dml_data,
ml_g, ml_m, ml_r,
n_folds,
ml_l, ml_m, ml_r, ml_g,
n_folds=n_folds,
score=score,
dml_procedure=dml_procedure)

dml_pliv_obj.fit()
Expand All @@ -51,7 +56,7 @@ def dml_pliv_pyvsr_fixture(generate_data_pliv, score, dml_procedure):
all_train, all_test = export_smpl_split_to_r(dml_pliv_obj.smpls[0])

r_dataframe = pandas2ri.py2rpy(obj_dml_data.data)
res_r = r_MLPLIV(r_dataframe, 'partialling out', dml_procedure,
res_r = r_MLPLIV(r_dataframe, score, dml_procedure,
all_train, all_test)
print(res_r)

Expand All @@ -76,22 +81,22 @@ def test_dml_pliv_pyvsr_se(dml_pliv_pyvsr_fixture):


@pytest.fixture(scope='module')
def dml_pliv_partial_x_pyvsr_fixture(generate_data_pliv_partialX, score, dml_procedure):
def dml_pliv_partial_x_pyvsr_fixture(generate_data_pliv_partialX, dml_procedure):
n_folds = 2

# collect data
obj_dml_data = generate_data_pliv_partialX

# Set machine learning methods for g, m & r
# Set machine learning methods for l, m & r
learner = LinearRegression()
ml_g = clone(learner)
ml_l = clone(learner)
ml_m = clone(learner)
ml_r = clone(learner)

np.random.seed(3141)
dml_pliv_obj = dml.DoubleMLPLIV(obj_dml_data,
ml_g, ml_m, ml_r,
n_folds,
ml_l, ml_m, ml_r,
n_folds=n_folds,
dml_procedure=dml_procedure)

dml_pliv_obj.fit()
Expand Down Expand Up @@ -125,20 +130,20 @@ def test_dml_pliv_partial_x_pyvsr_se(dml_pliv_partial_x_pyvsr_fixture):


@pytest.fixture(scope='module')
def dml_pliv_partial_z_pyvsr_fixture(generate_data_pliv_partialZ, score, dml_procedure):
def dml_pliv_partial_z_pyvsr_fixture(generate_data_pliv_partialZ, dml_procedure):
n_folds = 2

# collect data
obj_dml_data = generate_data_pliv_partialZ

# Set machine learning methods for g, m & r
# Set machine learning methods for r
learner = LinearRegression()
ml_r = clone(learner)

np.random.seed(3141)
dml_pliv_obj = dml.DoubleMLPLIV._partialZ(obj_dml_data,
ml_r,
n_folds,
n_folds=n_folds,
dml_procedure=dml_procedure)

dml_pliv_obj.fit()
Expand Down Expand Up @@ -172,22 +177,22 @@ def test_dml_pliv_partial_z_pyvsr_se(dml_pliv_partial_z_pyvsr_fixture):


@pytest.fixture(scope='module')
def dml_pliv_partial_xz_pyvsr_fixture(generate_data_pliv_partialXZ, score, dml_procedure):
def dml_pliv_partial_xz_pyvsr_fixture(generate_data_pliv_partialXZ, dml_procedure):
n_folds = 2

# collect data
obj_dml_data = generate_data_pliv_partialXZ

# Set machine learning methods for g, m & r
# Set machine learning methods for l, m & r
learner = LinearRegression()
ml_g = clone(learner)
ml_l = clone(learner)
ml_m = clone(learner)
ml_r = clone(learner)

np.random.seed(3141)
dml_pliv_obj = dml.DoubleMLPLIV._partialXZ(obj_dml_data,
ml_g, ml_m, ml_r,
n_folds,
ml_l, ml_m, ml_r,
n_folds=n_folds,
dml_procedure=dml_procedure)

dml_pliv_obj.fit()
Expand Down
12 changes: 8 additions & 4 deletions doubleml_py_vs_r/tests/test_plr_pyvsr.py
View file
Open in desktop
Original file line number Diff line number Diff line change
Expand Up @@ -37,14 +37,18 @@ def dml_plr_pyvsr_fixture(generate_data_plr, score, dml_procedure, n_rep):
# collect data
obj_dml_data = generate_data_plr

# Set machine learning methods for m & g
# Set machine learning methods for l, m & g
learner = LinearRegression()
ml_g = clone(learner)
ml_l = clone(learner)
ml_m = clone(learner)
if score == 'IV-type':
ml_g = clone(learner)
else:
ml_g = None

dml_plr_obj = dml.DoubleMLPLR(obj_dml_data,
ml_g, ml_m,
n_folds,
ml_l, ml_m, ml_g,
n_folds=n_folds,
n_rep=n_rep,
score=score,
dml_procedure=dml_procedure)
Expand Down