Merge pull request #28 from molpopgen/update_to_tskit_0_3_5

Update C files to tskit 0.3.5

Author: molpopgen

subprojects/tskit/tskit/core.c

@@ -281,8 +281,8 @@ tsk_strerror_internal(int err)
         case TSK_ERR_MUTATION_PARENT_AFTER_CHILD:
             ret = "Parent mutation ID must be < current ID";
             break;
-        case TSK_ERR_INCONSISTENT_MUTATIONS:
-            ret = "Inconsistent mutations: state already equal to derived state";
+        case TSK_ERR_MUTATION_PARENT_INCONSISTENT:
+            ret = "Mutation parent references form a loop.";
             break;
         case TSK_ERR_UNSORTED_MUTATIONS:
             ret = "Mutations must be provided in non-decreasing site order and "
@@ -357,10 +357,6 @@ tsk_strerror_internal(int err)
         case TSK_ERR_NONBINARY_MUTATIONS_UNSUPPORTED:
             ret = "Only binary mutations are supported for this operation";
             break;
-        case TSK_ERR_UNION_NOT_SUPPORTED:
-            ret = "Union is not supported for cases where there is non-shared"
-                  "history older than the shared history of the two Table Collections";
-            break;
 
         /* Stats errors */
         case TSK_ERR_BAD_NUM_WINDOWS:
@@ -470,6 +466,25 @@ tsk_strerror_internal(int err)
         case TSK_ERR_DUPLICATE_SAMPLE_PAIRS:
             ret = "There are duplicate sample pairs.";
             break;
+
+        /* Simplify errors */
+        case TSK_ERR_KEEP_UNARY_MUTUALLY_EXCLUSIVE:
+            ret = "You cannot specify both KEEP_UNARY and KEEP_UNARY_IN_INDIVDUALS.";
+            break;
+
+        /* Individual errors */
+        case TSK_ERR_UNSORTED_INDIVIDUALS:
+            ret = "Individuals must be provided in an order where children are after "
+                  "their parent individuals";
+            break;
+
+        case TSK_ERR_INDIVIDUAL_SELF_PARENT:
+            ret = "Individuals cannot be their own parents.";
+            break;
+
+        case TSK_ERR_INDIVIDUAL_PARENT_CYCLE:
+            ret = "Individuals cannot be their own ancestor.";
+            break;
     }
     return ret;
 }

subprojects/tskit/tskit/core.h

@@ -129,7 +129,7 @@ to the API or ABI are introduced, i.e., the addition of a new function.
 The library patch version. Incremented when any changes not relevant to the
 to the API or ABI are introduced, i.e., internal refactors of bugfixes.
 */
-#define TSK_VERSION_PATCH   9
+#define TSK_VERSION_PATCH   11
 /** @} */
 
 /* Node flags */
@@ -144,7 +144,7 @@ to the API or ABI are introduced, i.e., internal refactors of bugfixes.
 #define TSK_FILE_FORMAT_NAME          "tskit.trees"
 #define TSK_FILE_FORMAT_NAME_LENGTH   11
 #define TSK_FILE_FORMAT_VERSION_MAJOR 12
-#define TSK_FILE_FORMAT_VERSION_MINOR 3
+#define TSK_FILE_FORMAT_VERSION_MINOR 4
 
 /**
 @defgroup GENERAL_ERROR_GROUP General errors.
@@ -246,13 +246,13 @@ not found in the file.
 #define TSK_ERR_MUTATION_PARENT_DIFFERENT_SITE                      -500
 #define TSK_ERR_MUTATION_PARENT_EQUAL                               -501
 #define TSK_ERR_MUTATION_PARENT_AFTER_CHILD                         -502
-#define TSK_ERR_INCONSISTENT_MUTATIONS                              -503
-#define TSK_ERR_UNSORTED_MUTATIONS                                  -505
-#define TSK_ERR_NON_SINGLE_CHAR_MUTATION                            -506
-#define TSK_ERR_MUTATION_TIME_YOUNGER_THAN_NODE                     -507
-#define TSK_ERR_MUTATION_TIME_OLDER_THAN_PARENT_MUTATION            -508
-#define TSK_ERR_MUTATION_TIME_OLDER_THAN_PARENT_NODE                -509
-#define TSK_ERR_MUTATION_TIME_HAS_BOTH_KNOWN_AND_UNKNOWN            -510
+#define TSK_ERR_MUTATION_PARENT_INCONSISTENT                        -503
+#define TSK_ERR_UNSORTED_MUTATIONS                                  -504
+#define TSK_ERR_NON_SINGLE_CHAR_MUTATION                            -505
+#define TSK_ERR_MUTATION_TIME_YOUNGER_THAN_NODE                     -506
+#define TSK_ERR_MUTATION_TIME_OLDER_THAN_PARENT_MUTATION            -507
+#define TSK_ERR_MUTATION_TIME_OLDER_THAN_PARENT_NODE                -508
+#define TSK_ERR_MUTATION_TIME_HAS_BOTH_KNOWN_AND_UNKNOWN            -509
 
 /* Sample errors */
 #define TSK_ERR_DUPLICATE_SAMPLE                                    -600
@@ -274,7 +274,6 @@ not found in the file.
 #define TSK_ERR_SORT_OFFSET_NOT_SUPPORTED                           -803
 #define TSK_ERR_NONBINARY_MUTATIONS_UNSUPPORTED                     -804
 #define TSK_ERR_MIGRATIONS_NOT_SUPPORTED                            -805
-#define TSK_ERR_UNION_NOT_SUPPORTED                                 -806
 
 /* Stats errors */
 #define TSK_ERR_BAD_NUM_WINDOWS                                     -900
@@ -320,6 +319,14 @@ not found in the file.
 #define TSK_ERR_NO_SAMPLE_PAIRS                                    -1500
 #define TSK_ERR_DUPLICATE_SAMPLE_PAIRS                             -1501
 
+/* Simplify errors */
+#define TSK_ERR_KEEP_UNARY_MUTUALLY_EXCLUSIVE                      -1600
+
+/* Individual errors */
+#define TSK_ERR_UNSORTED_INDIVIDUALS                               -1700
+#define TSK_ERR_INDIVIDUAL_SELF_PARENT                             -1701
+#define TSK_ERR_INDIVIDUAL_PARENT_CYCLE                            -1702
+
 // clang-format on
 
 /* This bit is 0 for any errors originating from kastore */

subprojects/tskit/tskit/genotypes.c

@@ -159,6 +159,7 @@ tsk_vargen_init(tsk_vargen_t *self, const tsk_treeseq_t *tree_sequence,
     memset(self, 0, sizeof(tsk_vargen_t));
 
     if (samples == NULL) {
+        self->samples = tsk_treeseq_get_samples(tree_sequence);
         self->num_samples = tsk_treeseq_get_num_samples(tree_sequence);
         self->sample_index_map = tsk_treeseq_get_sample_index_map(tree_sequence);
         num_samples_alloc = self->num_samples;
@@ -171,6 +172,7 @@ tsk_vargen_init(tsk_vargen_t *self, const tsk_treeseq_t *tree_sequence,
         if (ret != 0) {
             goto out;
         }
+        self->samples = samples;
         self->num_samples = num_samples;
         self->sample_index_map = self->alt_sample_index_map;
     }
@@ -319,10 +321,7 @@ tsk_vargen_update_genotypes_i8_sample_list(
     if (index != TSK_NULL) {
         stop = list_right[node];
         while (true) {
-            if (genotypes[index] == (int8_t) derived) {
-                ret = TSK_ERR_INCONSISTENT_MUTATIONS;
-                goto out;
-            }
+
             ret += genotypes[index] == TSK_MISSING_DATA;
             genotypes[index] = (int8_t) derived;
             if (index == stop) {
@@ -331,7 +330,7 @@ tsk_vargen_update_genotypes_i8_sample_list(
             index = list_next[index];
         }
     }
-out:
+
     return ret;
 }
 
@@ -352,10 +351,7 @@ tsk_vargen_update_genotypes_i16_sample_list(
     if (index != TSK_NULL) {
         stop = list_right[node];
         while (true) {
-            if (genotypes[index] == (int16_t) derived) {
-                ret = TSK_ERR_INCONSISTENT_MUTATIONS;
-                goto out;
-            }
+
             ret += genotypes[index] == TSK_MISSING_DATA;
             genotypes[index] = (int16_t) derived;
             if (index == stop) {
@@ -364,7 +360,7 @@ tsk_vargen_update_genotypes_i16_sample_list(
             index = list_next[index];
         }
     }
-out:
+
     return ret;
 }
 
@@ -420,13 +416,10 @@ tsk_vargen_visit_i8(tsk_vargen_t *self, tsk_id_t sample_index, tsk_id_t derived)
 
     tsk_bug_assert(derived < INT8_MAX);
     tsk_bug_assert(sample_index != -1);
-    if (genotypes[sample_index] == (int8_t) derived) {
-        ret = TSK_ERR_INCONSISTENT_MUTATIONS;
-        goto out;
-    }
+
     ret = genotypes[sample_index] == TSK_MISSING_DATA;
     genotypes[sample_index] = (int8_t) derived;
-out:
+
     return ret;
 }
 
@@ -438,13 +431,10 @@ tsk_vargen_visit_i16(tsk_vargen_t *self, tsk_id_t sample_index, tsk_id_t derived
 
     tsk_bug_assert(derived < INT16_MAX);
     tsk_bug_assert(sample_index != -1);
-    if (genotypes[sample_index] == (int16_t) derived) {
-        ret = TSK_ERR_INCONSISTENT_MUTATIONS;
-        goto out;
-    }
+
     ret = genotypes[sample_index] == TSK_MISSING_DATA;
     genotypes[sample_index] = (int16_t) derived;
-out:
+
     return ret;
 }
 

subprojects/tskit/tskit/genotypes.h

@@ -52,9 +52,8 @@ typedef struct {
     size_t num_samples;
     size_t num_sites;
     const tsk_treeseq_t *tree_sequence;
-    const tsk_id_t *sample_index_map;
-    tsk_id_t *alt_samples;
-    tsk_id_t *alt_sample_index_map;
+    const tsk_id_t *samples;          /* samples being used */
+    const tsk_id_t *sample_index_map; /* reverse index map being used */
     bool user_alleles;
     char *user_alleles_mem;
     size_t tree_site_index;
@@ -63,6 +62,9 @@ typedef struct {
     tsk_tree_t tree;
     tsk_flags_t options;
     tsk_variant_t variant;
+    // private: the following data members are not intended to be used externally
+    tsk_id_t *alt_samples; /* alternative subset of samples to use */
+    tsk_id_t *alt_sample_index_map;
 } tsk_vargen_t;
 
 int tsk_vargen_init(tsk_vargen_t *self, const tsk_treeseq_t *tree_sequence,