!7012 [MD] sampler log update

Merge pull request !7012 from luoyang/pylint

mindspore/ccsrc/minddata/dataset/engine/datasetops/source/sampler/distributed_sampler.cc

@@ -76,7 +76,10 @@ Status DistributedSampler::InitSampler() {
 
 Status DistributedSampler::GetNextSample(std::unique_ptr<DataBuffer> *out_buffer) {
   if (cnt_ > samples_per_buffer_) {
-    RETURN_STATUS_UNEXPECTED("Distributed Sampler Error");
+    RETURN_STATUS_UNEXPECTED(
+      "Number of samples(cnt) that have already been filled in to buffer should be less than or "
+      "equal to samples_per_buffer, but got cnt: " +
+      std::to_string(cnt_) + ", samples_per_buffer: " + std::to_string(samples_per_buffer_));
   } else if (cnt_ == samples_per_buffer_ && (non_empty_ || !even_dist_)) {
     (*out_buffer) = std::make_unique<DataBuffer>(0, DataBuffer::kDeBFlagEOE);
     if (!samples_per_buffer_) {

mindspore/ccsrc/minddata/dataset/engine/datasetops/source/sampler/python_sampler.cc

@@ -53,7 +53,8 @@ Status PythonSampler::GetNextSample(std::unique_ptr<DataBuffer> *out_buffer) {
       } catch (const py::error_already_set &e) {
         return Status(StatusCode::kPyFuncException, e.what());
       } catch (const py::cast_error &e) {
-        return Status(StatusCode::kPyFuncException, "Python Sampler iterator should return integer index");
+        return Status(StatusCode::kPyFuncException,
+                      "Invalid data, python sampler iterator should return an integer index.");
       }
     }
     TensorRow row(1, sample_ids);
@@ -64,7 +65,8 @@ Status PythonSampler::GetNextSample(std::unique_ptr<DataBuffer> *out_buffer) {
 }
 
 Status PythonSampler::InitSampler() {
-  CHECK_FAIL_RETURN_UNEXPECTED(num_rows_ > 0, "ERROR num_rows_ should be greater than 0");
+  CHECK_FAIL_RETURN_UNEXPECTED(
+    num_rows_ > 0, "Invalid parameter, num_rows must be greater than 0, but got " + std::to_string(num_rows_));
   // Special value of 0 for num_samples means that the user wants to sample the entire set of data.
   // If the user asked to sample more rows than exists in the dataset, adjust the num_samples accordingly.
   if (num_samples_ == 0 || num_samples_ > num_rows_) {

mindspore/ccsrc/minddata/dataset/engine/datasetops/source/sampler/subset_random_sampler.cc

@@ -98,9 +98,8 @@ Status SubsetRandomSampler::GetNextSample(std::unique_ptr<DataBuffer> *out_buffe
     auto id_ptr = outputIds->begin<int64_t>();
     while (sample_id_ < last_id) {
       if (indices_[sample_id_] >= num_rows_) {
-        std::string err_msg =
-          "Generated id is bigger than numRows (out of bound). indices_: " + std::to_string(indices_[sample_id_]) +
-          " num_rows_: " + std::to_string(num_rows_);
+        std::string err_msg = "Generated indice is out of bound, expect range [0, num_data-1], got indice: " +
+                              std::to_string(indices_[sample_id_]) + ", num_data: " + std::to_string(num_rows_ - 1);
         RETURN_STATUS_UNEXPECTED(err_msg);
       }
 

mindspore/ccsrc/minddata/dataset/engine/datasetops/source/sampler/weighted_random_sampler.cc

@@ -50,11 +50,15 @@ Status WeightedRandomSampler::InitSampler() {
                                  std::to_string(samples_per_buffer_) + ".\n");
   if (weights_.size() > static_cast<size_t>(num_rows_)) {
     return Status(StatusCode::kUnexpectedError, __LINE__, __FILE__,
-                  "Invalid parameter, number of samples weights is more than num of rows. "
-                  "Might generate id out of bound OR other errors");
+                  "Invalid parameter, size of sample weights must be less than or equal to num of data, "
+                  "otherwise might cause generated id out of bound or other errors, but got weight size: " +
+                    std::to_string(weights_.size()) + ", num of data: " + std::to_string(num_rows_));
   }
   if (!replacement_ && (weights_.size() < static_cast<size_t>(num_samples_))) {
-    RETURN_STATUS_UNEXPECTED("Invalid parameter, without replacement, weights size must be greater than num_samples.");
+    RETURN_STATUS_UNEXPECTED(
+      "Invalid parameter, without replacement, weights size must be greater than or equal to num_samples, "
+      "but got weight size: " +
+      std::to_string(weights_.size()) + ", num_samples: " + std::to_string(num_samples_));
   }
 
   // Initialize random generator with seed from config manager
@@ -110,11 +114,16 @@ Status WeightedRandomSampler::ResetSampler() {
 Status WeightedRandomSampler::GetNextSample(std::unique_ptr<DataBuffer> *out_buffer) {
   if (weights_.size() > static_cast<size_t>(num_rows_)) {
     return Status(StatusCode::kUnexpectedError, __LINE__, __FILE__,
-                  "number of samples weights is more than num of rows. Might generate id out of bound OR other errors");
+                  "Invalid parameter, size of sample weights must be less than or equal to num of data, "
+                  "otherwise might cause generated id out of bound or other errors, but got weight size: " +
+                    std::to_string(weights_.size()) + ", num of data: " + std::to_string(num_rows_));
   }
 
   if (!replacement_ && (weights_.size() < static_cast<size_t>(num_samples_))) {
-    RETURN_STATUS_UNEXPECTED("Without replacement, sample weights less than numSamples");
+    RETURN_STATUS_UNEXPECTED(
+      "Invalid parameter, without replacement, weights size must be greater than or equal to num_samples, "
+      "but got weight size: " +
+      std::to_string(weights_.size()) + ", num_samples: " + std::to_string(num_samples_));
   }
 
   if (sample_id_ == num_samples_) {
@@ -150,7 +159,8 @@ Status WeightedRandomSampler::GetNextSample(std::unique_ptr<DataBuffer> *out_buf
       }
 
       if (genId >= num_rows_) {
-        RETURN_STATUS_UNEXPECTED("generated id is bigger than numRows (out of bound).");
+        RETURN_STATUS_UNEXPECTED("Generated indice is out of bound, expect range [0, num_data-1], got indice: " +
+                                 std::to_string(genId) + ", num_data: " + std::to_string(num_rows_ - 1));
       }
 
       if (HasChildSampler()) {