Add pydantic Field constraints

malariagen_data/anoph/base_params.py

@@ -2,6 +2,7 @@
 
 from typing import Final, List, Mapping, Optional, Sequence, Tuple, Union
 
+from pydantic import Field
 from typing_extensions import Annotated, TypeAlias
 
 from ..util import (
@@ -168,6 +169,7 @@ def _validate_sample_selection_params(
 
 cohort_size: TypeAlias = Annotated[
     int,
+    Field(ge=1),
     """
     Randomly down-sample to this value if the number of samples in the
     cohort is greater. Raise an error if the number of samples is less
@@ -177,6 +179,7 @@ def _validate_sample_selection_params(
 
 min_cohort_size: TypeAlias = Annotated[
     int,
+    Field(ge=1),
     """
     Minimum cohort size. Raise an error if the number of samples is
     less than this value.
@@ -185,6 +188,7 @@ def _validate_sample_selection_params(
 
 max_cohort_size: TypeAlias = Annotated[
     int,
+    Field(ge=1),
     """
     Randomly down-sample to this value if the number of samples in the
     cohort is greater.
@@ -193,6 +197,7 @@ def _validate_sample_selection_params(
 
 random_seed: TypeAlias = Annotated[
     int,
+    Field(ge=0),
     "Random seed used for reproducible down-sampling.",
 ]
 
@@ -228,6 +233,7 @@ def _validate_sample_selection_params(
 
 n_jack: TypeAlias = Annotated[
     int,
+    Field(ge=1),
     """
     Number of blocks to divide the data into for the block jackknife
     estimation of confidence intervals. N.B., larger is not necessarily
@@ -237,6 +243,7 @@ def _validate_sample_selection_params(
 
 confidence_level: TypeAlias = Annotated[
     float,
+    Field(gt=0, lt=1),
     """
     Confidence level to use for confidence interval calculation. E.g., 0.95
     means 95% confidence interval.
@@ -286,6 +293,7 @@ def _validate_sample_selection_params(
 
 n_snps: TypeAlias = Annotated[
     int,
+    Field(ge=1),
     """
     The desired number of SNPs to use when running the analysis.
     SNPs will be evenly thinned to approximately this number.
@@ -294,6 +302,7 @@ def _validate_sample_selection_params(
 
 thin_offset: TypeAlias = Annotated[
     int,
+    Field(ge=0),
     """
     Starting index for SNP thinning. Change this to repeat the analysis
     using a different set of SNPs.

malariagen_data/anoph/cnv_params.py

@@ -2,10 +2,12 @@
 
 from typing import Optional
 
+from pydantic import Field
 from typing_extensions import Annotated, TypeAlias
 
 max_coverage_variance: TypeAlias = Annotated[
     Optional[float],
+    Field(gt=0),
     """
     Remove samples if coverage variance exceeds this value.
     """,

malariagen_data/anoph/dash_params.py

@@ -2,9 +2,12 @@
 
 from typing import Literal, Union
 
+from pydantic import Field
 from typing_extensions import Annotated, TypeAlias
 
-height: TypeAlias = Annotated[int, "Height of the Dash app in pixels (px)."]
+height: TypeAlias = Annotated[
+    int, Field(gt=0), "Height of the Dash app in pixels (px)."
+]
 
 width: TypeAlias = Annotated[Union[int, str], "Width of the Dash app."]
 
@@ -21,6 +24,7 @@
 
 server_port: TypeAlias = Annotated[
     int,
+    Field(gt=0),
     "Manually override the port on which the Dash app will run.",
 ]
 

malariagen_data/anoph/fst_params.py

@@ -3,13 +3,15 @@
 from typing import Optional, Literal
 
 import pandas as pd
+from pydantic import Field
 from typing_extensions import Annotated, TypeAlias
 
 from . import base_params
 
 # N.B., window size can mean different things for different functions
 window_size: TypeAlias = Annotated[
     int,
+    Field(ge=1),
     "The size of windows (number of sites) used to calculate statistics within.",
 ]
 cohort_size_default: Optional[base_params.cohort_size] = None

malariagen_data/anoph/g123_params.py

@@ -2,6 +2,7 @@
 
 from typing import Sequence
 
+from pydantic import Field
 from typing_extensions import Annotated, TypeAlias
 
 from . import base_params
@@ -29,6 +30,7 @@
 
 window_size: TypeAlias = Annotated[
     int,
+    Field(ge=1),
     """
     The size of windows (number of sites) used to calculate statistics within.
     """,

malariagen_data/anoph/gplt_params.py

@@ -4,6 +4,7 @@
 from typing import Literal, Mapping, Optional, Union, Final, Sequence
 
 import bokeh.models
+from pydantic import Field
 from typing_extensions import Annotated, TypeAlias
 
 sizing_mode: TypeAlias = Annotated[
@@ -33,21 +34,25 @@
 
 height: TypeAlias = Annotated[
     int,
+    Field(gt=0),
     "Plot height in pixels (px).",
 ]
 
 row_height: TypeAlias = Annotated[
     int,
+    Field(gt=0),
     "Plot height per row (sample) in pixels (px).",
 ]
 
 track_height: TypeAlias = Annotated[
     int,
+    Field(gt=0),
     "Main track height in pixels (px).",
 ]
 
 genes_height: TypeAlias = Annotated[
     int,
+    Field(gt=0),
     "Genes track height in pixels (px).",
 ]
 

malariagen_data/anoph/h12_params.py

@@ -2,6 +2,7 @@
 
 from typing import Optional, Sequence, Union
 
+from pydantic import Field
 from typing_extensions import Annotated, TypeAlias
 
 from . import base_params
@@ -17,6 +18,7 @@
 
 window_size: TypeAlias = Annotated[
     int,
+    Field(ge=1),
     """
     The size of windows (number of SNPs) used to calculate statistics within.
     """,

malariagen_data/anoph/hapnet_params.py

@@ -2,17 +2,20 @@
 
 from typing import Mapping
 
+from pydantic import Field
 from typing_extensions import Annotated, TypeAlias
 
 max_dist: TypeAlias = Annotated[
     int,
+    Field(ge=0),
     "Join network components up to a maximum distance of 2 SNP differences.",
 ]
 
 max_dist_default: max_dist = 2
 
 node_size_factor: TypeAlias = Annotated[
     int,
+    Field(gt=0),
     "Control the sizing of nodes.",
 ]
 

malariagen_data/anoph/het_params.py

@@ -3,17 +3,20 @@
 from typing import Tuple
 
 import pandas as pd
+from pydantic import Field
 from typing_extensions import Annotated, TypeAlias
 
 window_size: TypeAlias = Annotated[
     int,
+    Field(ge=1),
     "Number of sites per window.",
 ]
 
 window_size_default: window_size = 20_000
 
 phet_roh: TypeAlias = Annotated[
     float,
+    Field(ge=0, le=1),
     "Probability of observing a heterozygote in a ROH.",
 ]
 
@@ -28,6 +31,7 @@
 
 transition: TypeAlias = Annotated[
     float,
+    Field(ge=0, le=1),
     """
     Probability of moving between states. A larger window size may call
     for a larger transitional probability.
@@ -38,6 +42,7 @@
 
 y_max: TypeAlias = Annotated[
     float,
+    Field(gt=0),
     "Y axis limit.",
 ]
 

malariagen_data/anoph/ihs_params.py

@@ -2,12 +2,14 @@
 
 from typing import Tuple, Union
 
+from pydantic import Field
 from typing_extensions import Annotated, TypeAlias
 
 from . import base_params
 
 window_size: TypeAlias = Annotated[
     int,
+    Field(ge=1),
     """
     The size of window in number of SNPs used to summarise iHS over.
     If None, per-variant iHS values are returned.
@@ -41,6 +43,7 @@
 
 standardization_n_bins: TypeAlias = Annotated[
     int,
+    Field(ge=1),
     """
     Number of allele count bins to use for standardization.
     Overrides standardization_bins.
@@ -55,6 +58,7 @@
 
 filter_min_maf: TypeAlias = Annotated[
     float,
+    Field(ge=0, le=0.5),
     """
     Minimum minor allele frequency to use for filtering prior to passing
     haplotypes to allel.ihs function
@@ -65,6 +69,7 @@
 
 compute_min_maf: TypeAlias = Annotated[
     float,
+    Field(ge=0, le=0.5),
     """
     Do not compute integrated haplotype homozygosity for variants with
     minor allele frequency below this threshold.
@@ -75,6 +80,7 @@
 
 min_ehh: TypeAlias = Annotated[
     float,
+    Field(ge=0, le=1),
     """
     Minimum EHH beyond which to truncate integrated haplotype homozygosity
     calculation.
@@ -85,6 +91,7 @@
 
 max_gap: TypeAlias = Annotated[
     int,
+    Field(ge=1),
     """
     Do not report scores if EHH spans a gap larger than this number of
     base pairs.
@@ -94,7 +101,9 @@
 max_gap_default: max_gap = 200_000
 
 gap_scale: TypeAlias = Annotated[
-    int, "Rescale distance between variants if gap is larger than this value."
+    int,
+    Field(ge=1),
+    "Rescale distance between variants if gap is larger than this value.",
 ]
 
 gap_scale_default: gap_scale = 20_000

malariagen_data/anoph/pca_params.py

@@ -2,11 +2,13 @@
 
 import numpy as np
 import pandas as pd
+from pydantic import Field
 from typing_extensions import Annotated, TypeAlias
 from . import base_params
 
 n_components: TypeAlias = Annotated[
     int,
+    Field(ge=1),
     "Number of components to return.",
 ]
 

malariagen_data/anoph/plotly_params.py

@@ -7,6 +7,7 @@
 from typing import List, Literal, Mapping, Optional, Union
 
 import plotly.graph_objects as go  # type: ignore
+from pydantic import Field
 from typing_extensions import Annotated, TypeAlias
 
 x_label: TypeAlias = Annotated[
@@ -118,11 +119,13 @@
 
 jitter_frac: TypeAlias = Annotated[
     Optional[float],
+    Field(ge=0, le=1),
     "Randomly jitter points by this fraction of their range.",
 ]
 
 marker_size: TypeAlias = Annotated[
     Union[int, float],
+    Field(gt=0),
     "Marker size.",
 ]
 

malariagen_data/anoph/xpehh_params.py

@@ -3,6 +3,7 @@
 from typing import Tuple, Union, Optional
 
 import numpy as np
+from pydantic import Field
 from typing_extensions import Annotated, TypeAlias
 
 from . import base_params
@@ -27,6 +28,7 @@
 percentiles_default: percentiles = (50, 75, 100)
 filter_min_maf: TypeAlias = Annotated[
     float,
+    Field(ge=0, le=0.5),
     """
     Minimum minor allele frequency to use for filtering prior to passing
     haplotypes to allel.xpehh function
@@ -41,6 +43,7 @@
 ]
 min_ehh: TypeAlias = Annotated[
     float,
+    Field(ge=0, le=1),
     """
     Minimum EHH beyond which to truncate integrated haplotype homozygosity
     calculation.
@@ -49,14 +52,17 @@
 min_ehh_default: min_ehh = 0.05
 max_gap: TypeAlias = Annotated[
     int,
+    Field(ge=1),
     """
     Do not report scores if EHH spans a gap larger than this number of
     base pairs.
     """,
 ]
 max_gap_default: max_gap = 200_000
 gap_scale: TypeAlias = Annotated[
-    int, "Rescale distance between variants if gap is larger than this value."
+    int,
+    Field(ge=1),
+    "Rescale distance between variants if gap is larger than this value.",
 ]
 gap_scale_default: gap_scale = 20_000
 include_edges: TypeAlias = Annotated[