hjxdl 0.1.12__py3-none-any.whl → 0.1.14__py3-none-any.whl

hdl/jupyfuncs/show/plot.py

@@ -0,0 +1,259 @@
+from os import path as osp
+import typing as t
+from typing_extensions import Literal
+
+import seaborn as sn
+import sklearn
+import matplotlib.pyplot as plt
+import matplotlib
+import numpy as np
+import pandas as pd
+
+from ..path.glob import get_num_lines
+from ..path.strings import splitted_strs_from_line 
+
+cm = matplotlib.cm.get_cmap('tab20')
+colors = cm.colors
+LABEL = Literal[
+    'training_size',
+    'episode_id',
+]
+
+
+def accuracies_heat(y_true, y_pred, num_tasks):
+    assert len(y_true) == len(y_pred)
+    cm = sklearn.metrics.confusion_matrix(
+        y_true, y_pred, normalize='true'
+    )
+    df_cm = pd.DataFrame(
+        cm, range(num_tasks), range(num_tasks)
+    )
+    plt.figure(figsize=(10, 10))
+    sn.set(font_scale=1.4) 
+    sn.heatmap(df_cm, annot=True, annot_kws={"size": 16}, fmt='.2f')
+
+
+def get_metrics_curves(
+    base_dir,
+    ckpts,
+    num_points,
+    title="Metric Curve",
+    metric='accuracy',
+    log_file='metrics.log',
+    label: LABEL = 'training_size',
+    save_dir: str = None,
+    figsize=(10, 6)
+):
+    if not save_dir:
+        save_dir = osp.join(base_dir, 'metrics_curves.png')
+    data_dict = {}
+    for ckpt in ckpts:
+        log = osp.join(
+            base_dir,
+            ckpt,
+            log_file
+        )
+        if not osp.exists(log):
+            print(f"WARNING: no log file for {ckpt}")
+            continue
+        data_dict[ckpt] = []
+        data_idx = 0
+        for line_id in range(get_num_lines(log)):
+            line = splitted_strs_from_line(log, line_id)
+            if len(line) == 3 and line[1].strip() == metric:
+                if label == 'episode_id':
+                    x = data_idx
+                elif label == 'training_size':
+                    x = int(line[0].strip())
+                data_dict[ckpt].append(
+                    [
+                        x,
+                        float(line[2].strip())
+                    ]
+                )
+                data_idx += 1
+            if line_id >= num_points - 1:
+                break
+    plt.figure(figsize=figsize, dpi=100)
+    # plt.style.use('ggplot')
+    plt.title(title)
+    for i, (ckpt, points) in enumerate(data_dict.items()):
+        points_array = np.array(points).T
+        plt.plot(points_array[0], points_array[1], label=ckpt, color=colors[i])
+    lg = plt.legend(bbox_to_anchor=(1.2, 1.0), loc='upper right')
+    # plt.legend(loc='lower right')
+    plt.xlabel(
+        label
+    )
+    plt.ylabel(metric)
+    plt.grid(True)
+    plt.savefig(
+        save_dir,
+        format='png', 
+        bbox_extra_artists=(lg,), 
+        bbox_inches='tight'
+    )
+    plt.show()
+
+
+def get_means_vars(
+    log_file: str,
+    indices: t.List,
+    mode: str,
+    nears_each: int,
+) -> t.List[t.List[int]]:
+    
+    mean_s, var_s = [], []
+    num_points = len(indices)
+
+    nears_lists = []
+    if mode == 'id':
+
+        for index in indices:
+            nears = []
+            nears.extend(list(range(
+                index - nears_each, index + 1 + nears_each
+            ))) 
+            nears_lists.append(nears)
+        
+        for nears in nears_lists:
+            mean_s.append(np.mean([
+                float(splitted_strs_from_line(log_file, line_id)[2])
+                for line_id in nears
+            ]))
+            var_s.append((np.std([
+                float(splitted_strs_from_line(log_file, line_id)[2])
+                for line_id in nears
+            ])))
+
+    elif mode == 'value':
+        datas = [
+            splitted_strs_from_line(log_file, line_id)
+            for line_id in range(get_num_lines(log_file))
+        ]
+        training_sizes = [[int(data[0]) for data in datas]]
+        values = np.array([float(data[2]) for data in datas])
+
+        training_sizes = np.repeat(training_sizes, num_points, 0).T
+        diffs = training_sizes - indices
+
+        true_indices = np.argmin(np.abs(diffs), 0)
+
+        true_indices_list = [
+            list(range(
+                index - nears_each, index + 1 + nears_each
+            ))
+            for index in true_indices
+        ]
+        mean_s = [
+            np.mean(values[indices])
+            for indices in true_indices_list
+        ]
+        var_s = [
+            np.std(values[indices])
+            for indices in true_indices_list
+        ]
+        var_s = np.array(var_s) / np.sqrt(num_points)
+ 
+    return mean_s, var_s 
+
+
+def get_metrics_bars(
+    base_dir,
+    ckpts,
+    title="Metric Bars",
+    training_sizes: t.List = [],
+    episide_ids: t.List = [],
+    nears_each: int = 5,
+    pretrained_num: int = 0,
+    x_diff: bool = False,
+    metric='accuracy',
+    log_file='metrics.log',
+    label: LABEL = 'training_size',
+    save_dir: str = None,
+    figsize=(10, 6),
+    bar_ratio=0.8,
+    minimum=0.0,
+    maximum=1.0
+):
+
+    if not save_dir:
+        save_dir = osp.join(base_dir, 'metrics_bars.png')
+    
+    x_labels, num_points = [], 0
+    if label == 'training_size':
+        num_points = len(training_sizes)
+        x_labels = training_sizes
+        mode = 'value'
+    elif label == 'episode_id':
+        num_points = len(episide_ids)
+        x_labels = episide_ids
+        mode = 'id'
+    
+    x = np.arange(num_points)
+    num_strategies = len(ckpts)
+    total_width = bar_ratio
+    width = total_width / num_strategies  
+    x = x - (total_width - width) / 2
+        
+    if not save_dir:
+        save_dir = osp.join(base_dir, 'metrics.png')
+
+    data_dict = {}
+    for ckpt in ckpts:
+
+        log = osp.join(
+            base_dir,
+            ckpt,
+            log_file
+        )
+        if not osp.exists(log):
+            print(f"WARNING: no log file for {ckpt}")
+            continue
+        
+        # print(x_labels)
+        mean_s, var_s = get_means_vars(
+            log_file=log,
+            indices=x_labels,
+            mode=mode,
+            nears_each=nears_each
+        )
+        data_dict[ckpt] = (mean_s, var_s)
+    
+    if x_diff:
+        x_labels = np.array(x_labels, dtype=np.int) - pretrained_num
+ 
+    plt.figure(figsize=figsize, dpi=100)
+    plt.title(title)
+    ax = plt.gca()
+    ax.set_ylim([minimum, maximum])
+    
+    for point_idx, (ckpt, datas) in enumerate(data_dict.items()):
+        mean_s, var_s = datas
+        plt.bar(
+            x + width * point_idx,
+            mean_s, 
+            width=width,
+            yerr=var_s,
+            tick_label=x_labels,
+            label=ckpt,
+            color=colors[point_idx]
+        )
+    
+    lg = plt.legend(bbox_to_anchor=(1.2, 1.0), loc='upper right')
+    # plt.legend(loc='lower right')
+    plt.xlabel(
+        label
+    )
+    plt.ylabel(metric)
+    plt.grid(True)
+ 
+    plt.savefig(
+        save_dir,
+        format='png', 
+        bbox_extra_artists=(lg,), 
+        bbox_inches='tight'
+    )
+    plt.show()
+
+

hdl/jupyfuncs/utils/wrappers.py

@@ -0,0 +1,8 @@
+class GeneratorWrapper:
+    def __init__(self, generator_func, *args, **kwargs):
+        self.generator_func = generator_func
+        self.args = args
+        self.kwargs = kwargs
+
+    def __iter__(self):
+        return self.generator_func(*self.args, **self.kwargs)

hdl/utils/llm/chat.py

@@ -364,6 +364,7 @@ class OpenAI_M():
         prompt : str,
         stop: list[str] | None = ["USER:", "ASSISTANT:"],
         # history: list = [],
+        model="default_model",
         **kwargs: t.Any,
     ) -> str:
         """Invoke the chatbot with the given prompt and return the response.
@@ -383,6 +384,7 @@ class OpenAI_M():
                 "content": prompt
             }],
             stream=False,
+            model=model,
             # model=kwargs.get("model", "default_model")
             **kwargs
         )
@@ -393,6 +395,7 @@ class OpenAI_M():
         prompt : str,
         stop: list[str] | None = ["USER:", "ASSISTANT:"],
         # history: list = [],
+        model="default_model",
         **kwargs: t.Any,
     ):
         """Generate text completion in a streaming fashion.
@@ -412,6 +415,7 @@ class OpenAI_M():
                 "content": prompt
             }],
             stream=True,
+            model=model,
             # model=kwargs.get("model", "default_model")
             **kwargs
         )

{hjxdl-0.1.12.dist-info → hjxdl-0.1.14.dist-info}/METADATA

@@ -1,6 +1,6 @@
 Metadata-Version: 2.1
 Name: hjxdl
-Version: 0.1.12
+Version: 0.1.14
 Summary: A collection of functions for Jupyter notebooks
 Home-page: https://github.com/huluxiaohuowa/hdl
 Author: Jianxing Hu

{hjxdl-0.1.12.dist-info → hjxdl-0.1.14.dist-info}/RECORD

@@ -1,5 +1,5 @@
 hdl/__init__.py,sha256=5sZZNySv08wwfzJcSDssGTqUn9wlmDsR6R4XB8J8mFM,70
-hdl/_version.py,sha256=cuLjhzc4k2qL9vS7TDaPLs82wisSY2yFKQtDhC08Kis,413
+hdl/_version.py,sha256=1w8OANeC0K0uOMBB6XWDhN49spC7ipRwXZnxjNMUZWU,413
 hdl/args/__init__.py,sha256=47DEQpj8HBSa-_TImW-5JCeuQeRkm5NMpJWZG3hSuFU,0
 hdl/args/loss_args.py,sha256=s7YzSdd7IjD24rZvvOrxLLFqMZQb9YylxKeyelSdrTk,70
 hdl/controllers/__init__.py,sha256=47DEQpj8HBSa-_TImW-5JCeuQeRkm5NMpJWZG3hSuFU,0
@@ -38,6 +38,11 @@ hdl/data/dataset/samplers/__init__.py,sha256=47DEQpj8HBSa-_TImW-5JCeuQeRkm5NMpJW
 hdl/data/dataset/samplers/chiral.py,sha256=ZS83kg5e2gdHVGgIuCjCepDwk2SKqWDgJawH37oXy78,463
 hdl/data/dataset/seq/__init__.py,sha256=47DEQpj8HBSa-_TImW-5JCeuQeRkm5NMpJWZG3hSuFU,0
 hdl/data/dataset/seq/rxn_dataset.py,sha256=jfXFlR3ITAf0KwUfIevzUZHnLBnFYrL69Cc81EMv0x0,1668
+hdl/datasets/city_code.json,sha256=qnTL6ldpGnQanDXN3oitx12E6oiayaCHTh2Zi9RyQjM,60816
+hdl/datasets/defined_BaseFeatures.fdef,sha256=5QhCEcu6fjSTXaTcZZ8-LSgf72_aJj_ykoDk82ZwVBI,7383
+hdl/datasets/las.tsv,sha256=47DEQpj8HBSa-_TImW-5JCeuQeRkm5NMpJWZG3hSuFU,0
+hdl/datasets/route_template.json,sha256=2qhkbtEZUrUod6PXCWXxAgQmU-jAC0yLcWGBBk2IwgE,3757
+hdl/datasets/vocab.txt,sha256=cXdB1JDs2rY6C9BupRRY7w21xK4KN9SrxPxkN9CUXXQ,3524
 hdl/features/__init__.py,sha256=47DEQpj8HBSa-_TImW-5JCeuQeRkm5NMpJWZG3hSuFU,0
 hdl/features/fp/__init__.py,sha256=47DEQpj8HBSa-_TImW-5JCeuQeRkm5NMpJWZG3hSuFU,0
 hdl/features/fp/features_generators.py,sha256=HbyS97i2I2mOcANdJMohs2okA1LlZmkG4ZIIX6Y9fr4,9017
@@ -45,6 +50,40 @@ hdl/features/graph/__init__.py,sha256=47DEQpj8HBSa-_TImW-5JCeuQeRkm5NMpJWZG3hSuF
 hdl/features/graph/featurization.py,sha256=QLbj33JsgO-OWarIC2HXQP7eMu8pd-GWmppZQj_tQ_k,10902
 hdl/features/utils/__init__.py,sha256=47DEQpj8HBSa-_TImW-5JCeuQeRkm5NMpJWZG3hSuFU,0
 hdl/features/utils/utils.py,sha256=aL4UAALblaw1w0AjK7MX8nSj9zwTmrp9CTLwJUX8ZtE,4225
+hdl/ju/__init__.py,sha256=47DEQpj8HBSa-_TImW-5JCeuQeRkm5NMpJWZG3hSuFU,0
+hdl/ju/setup.py,sha256=MB3rndQYt9QC-bdyGt81HYR0Rdr0l8DbAHktIuFMYU0,1725
+hdl/jupyfuncs/__init__.py,sha256=47DEQpj8HBSa-_TImW-5JCeuQeRkm5NMpJWZG3hSuFU,0
+hdl/jupyfuncs/chem/__init__.py,sha256=47DEQpj8HBSa-_TImW-5JCeuQeRkm5NMpJWZG3hSuFU,0
+hdl/jupyfuncs/chem/mol.py,sha256=FDb2L61JL6xkNK7bxXWVjAT-r-st7iMyQNoFezBvTUE,15294
+hdl/jupyfuncs/chem/norm.py,sha256=NRjSP8P7f3Yhy1LmSpaV93qYgYis_qVG7HT0vOWeW1U,10186
+hdl/jupyfuncs/chem/pdb_ext.py,sha256=VgkU34Y3na5Ri0QwW2Hh-QiE4j9DgVwMB49DL3JgAcQ,2710
+hdl/jupyfuncs/chem/scaffold.py,sha256=hWhmsrtdjM_ihStFvTz7XIs27coOMDOvdGh5vAk-UsA,1025
+hdl/jupyfuncs/chem/shape.py,sha256=FQmLxEkUOaWqlYi1CPm-8Ze5w_oOrmoeAzOx4UdYRtI,5807
+hdl/jupyfuncs/chem/tokenizers.py,sha256=UtpgQdDCL3FF1gSYcMLWZZcwcJQSp2RXdG0fRYZDGmU,133
+hdl/jupyfuncs/dbtools/__init__.py,sha256=47DEQpj8HBSa-_TImW-5JCeuQeRkm5NMpJWZG3hSuFU,0
+hdl/jupyfuncs/dbtools/pg.py,sha256=B4t1WAfGCmJsaQIm59M5xYdRPfGP-vrNS2SA8rMzlbM,943
+hdl/jupyfuncs/dbtools/query_info.py,sha256=CorUpyopY-FO3mRXlCIjxD09_6nXaAMJR0NHvfqjcIw,4327
+hdl/jupyfuncs/dl/__init__.py,sha256=47DEQpj8HBSa-_TImW-5JCeuQeRkm5NMpJWZG3hSuFU,0
+hdl/jupyfuncs/dl/cp.py,sha256=NHBkWBQ0eP6qEUXMwsVv5bI0p24CEsqFhwSJtKmmzGo,1816
+hdl/jupyfuncs/dl/dataframe.py,sha256=Cqlkcv20QplZ4tGQdW9k-3CaoTaaqoyaWQ0mr96n3V0,1023
+hdl/jupyfuncs/dl/fp.py,sha256=VYiW6-Q8lY2fDK9DU5_r6EmAyY5-VVv680-57h_MKPA,1025
+hdl/jupyfuncs/dl/list.py,sha256=WcVoDoJU7LYAG0bp_3zcZF2pW9NfQ4LIIXvGGu32vzs,756
+hdl/jupyfuncs/dl/model_utils.py,sha256=Xnt-a5MjE5NWghIu9Yss8zQ9JYhrv_Rbo5OPNC0MX80,2934
+hdl/jupyfuncs/dl/tensor.py,sha256=vOUyi8Ymc_JKiqywLJUph5jiKPHoq2cUo73YJnf0voQ,4495
+hdl/jupyfuncs/dl/uncs.py,sha256=j4YQE4BrBmZqOk9J_C9UKxbzrvt5pLPUZCCSIdi2FBk,2780
+hdl/jupyfuncs/llm/__init__.py,sha256=47DEQpj8HBSa-_TImW-5JCeuQeRkm5NMpJWZG3hSuFU,0
+hdl/jupyfuncs/llm/extract.py,sha256=UjPhrgbuU7_2u3620lcjXQPHF2l22o_AEmAaH8UXIZA,4531
+hdl/jupyfuncs/llm/openapi.py,sha256=pNBW0Jzt0JAZP8ZexgoQZVF118jrhKnBYheClcMu9bU,2687
+hdl/jupyfuncs/network/__init__.py,sha256=47DEQpj8HBSa-_TImW-5JCeuQeRkm5NMpJWZG3hSuFU,0
+hdl/jupyfuncs/network/proxy.py,sha256=foZm3gGFTPLeMNRfWs4QKNUEmlhtNTr_1GQvn0IgDVw,545
+hdl/jupyfuncs/path/__init__.py,sha256=47DEQpj8HBSa-_TImW-5JCeuQeRkm5NMpJWZG3hSuFU,0
+hdl/jupyfuncs/path/glob.py,sha256=fl0YDLDS9QI2WgDlBzDGlALbvkmPGcXp4UnbbQ17BOM,8300
+hdl/jupyfuncs/path/strings.py,sha256=eZCXElh7pT0xwy6ZBqSu3frq3Xx8CN5TMuQzsxb0Sbw,2009
+hdl/jupyfuncs/show/__init__.py,sha256=47DEQpj8HBSa-_TImW-5JCeuQeRkm5NMpJWZG3hSuFU,0
+hdl/jupyfuncs/show/pbar.py,sha256=QzyHV9XEyk9U5oyonxLSwYb5pD09iVUw_atVlzxYBNQ,1005
+hdl/jupyfuncs/show/plot.py,sha256=kH4UwTiRtjhTJJlR4gj6-mDgQxIH9PQCHVg1_54nIR8,6717
+hdl/jupyfuncs/utils/__init__.py,sha256=47DEQpj8HBSa-_TImW-5JCeuQeRkm5NMpJWZG3hSuFU,0
+hdl/jupyfuncs/utils/wrappers.py,sha256=N8rwRFnHX9j2sR_-l-y-B2nmk39NFuu4YTntOvF3pMo,266
 hdl/layers/__init__.py,sha256=47DEQpj8HBSa-_TImW-5JCeuQeRkm5NMpJWZG3hSuFU,0
 hdl/layers/general/__init__.py,sha256=47DEQpj8HBSa-_TImW-5JCeuQeRkm5NMpJWZG3hSuFU,0
 hdl/layers/general/gp.py,sha256=no1P6i2nCa539b0I5S6hd2mC8CeW0Ds726GM0swwlzc,525
@@ -84,12 +123,12 @@ hdl/utils/database_tools/connect.py,sha256=KUnVG-8raifEJ_N0b3c8LkTTIfn9NIyw8LX6q
 hdl/utils/general/__init__.py,sha256=47DEQpj8HBSa-_TImW-5JCeuQeRkm5NMpJWZG3hSuFU,0
 hdl/utils/general/glob.py,sha256=8-RCnt6L297wMIfn34ZAMCsGCZUjHG3MGglGZI1cX0g,491
 hdl/utils/llm/__init__.py,sha256=47DEQpj8HBSa-_TImW-5JCeuQeRkm5NMpJWZG3hSuFU,0
-hdl/utils/llm/chat.py,sha256=FtpQjp87FIBSIOQBPYKjtscA13RpMRwLk7sKlMAnOQ0,14974
+hdl/utils/llm/chat.py,sha256=Xypu7KlnTeI9oep02tuw04GooAfysZEgY2T_P29xZuU,15086
 hdl/utils/llm/embs.py,sha256=Tf0FOYrOFZp7qQpEPiSCXzlgyHH0X9HVTUtsup74a9E,7174
 hdl/utils/llm/extract.py,sha256=2sK_WJzmYIc8iuWaM9DA6Nw3_6q1O4lJ5pKpcZo-bBA,6512
 hdl/utils/schedulers/__init__.py,sha256=47DEQpj8HBSa-_TImW-5JCeuQeRkm5NMpJWZG3hSuFU,0
 hdl/utils/schedulers/norm_lr.py,sha256=bDwCmdEK-WkgxQMFBiMuchv8Mm7C0-GZJ6usm-PQk14,4461
-hjxdl-0.1.12.dist-info/METADATA,sha256=oM9y6W0hDN6TupHYR2OECwnI1sI7BLlpWqkbOV9slK0,543
-hjxdl-0.1.12.dist-info/WHEEL,sha256=R0nc6qTxuoLk7ShA2_Y-UWkN8ZdfDBG2B6Eqpz2WXbs,91
-hjxdl-0.1.12.dist-info/top_level.txt,sha256=-kxwTM5JPhylp06z3zAVO3w6_h7wtBfBo2zgM6YZoTk,4
-hjxdl-0.1.12.dist-info/RECORD,,
+hjxdl-0.1.14.dist-info/METADATA,sha256=zSqAbKEZutZyv5WQguUpas0Jm_nTHdBsdUyBpT6o8lQ,543
+hjxdl-0.1.14.dist-info/WHEEL,sha256=Mdi9PDNwEZptOjTlUcAth7XJDFtKrHYaQMPulZeBCiQ,91
+hjxdl-0.1.14.dist-info/top_level.txt,sha256=-kxwTM5JPhylp06z3zAVO3w6_h7wtBfBo2zgM6YZoTk,4
+hjxdl-0.1.14.dist-info/RECORD,,

{hjxdl-0.1.12.dist-info → hjxdl-0.1.14.dist-info}/WHEEL

@@ -1,5 +1,5 @@
 Wheel-Version: 1.0
-Generator: setuptools (72.1.0)
+Generator: setuptools (73.0.1)
 Root-Is-Purelib: true
 Tag: py3-none-any