@nshiab/simple-data-analysis 5.19.8

package/esm/deps/jsr.io/@nshiab/simple-data-analysis-core/0.0.17/src/methods/summarizeQuery.d.ts.map

@@ -0,0 +1 @@
+{"version":3,"file":"summarizeQuery.d.ts","sourceRoot":"","sources":["../../../../../../../../src/deps/jsr.io/@nshiab/simple-data-analysis-core/0.0.17/src/methods/summarizeQuery.ts"],"names":[],"mappings":"AAAA,MAAM,CAAC,OAAO,UAAU,cAAc,CACpC,KAAK,EAAE,MAAM,EACb,KAAK,EAAE;IACL,CAAC,GAAG,EAAE,MAAM,GAAG,MAAM,CAAC;CACvB,EACD,WAAW,EAAE,MAAM,EACnB,MAAM,EAAE,MAAM,EAAE,EAChB,UAAU,EAAE,MAAM,EAAE,EACpB,SAAS,EAAE,CACP,OAAO,GACP,aAAa,GACb,WAAW,GACX,KAAK,GACL,KAAK,GACL,MAAM,GACN,QAAQ,GACR,KAAK,GACL,MAAM,GACN,QAAQ,GACR,KAAK,CACR,EAAE,EACH,OAAO,EAAE;IAAE,QAAQ,CAAC,EAAE,MAAM,CAAC;IAAC,aAAa,CAAC,EAAE,OAAO,CAAA;CAAE,YAAK,EAC5D,OAAO,EAAE,MAAM,EAAE,GAAG,SAAS,UA4J9B"}

package/esm/deps/jsr.io/@nshiab/simple-data-analysis-core/0.0.17/src/methods/summarizeQuery.js

@@ -0,0 +1,112 @@
+export default function summarizeQuery(table, types, outputTable, values, categories, summaries, options = {}, columns) {
+    if (values.length > 1 && options.noColumnValue) {
+        throw new Error("The options `noColumnValue` works only when you aggregate the values from one column. Remove the option `noColumnValue` or specify just one column in the option `values`.");
+    }
+    const typesOfValues = values.map((d) => types[d]);
+    const doubleAndDate = Object.values(typesOfValues).includes("DOUBLE") &&
+        Object.values(typesOfValues).filter((d) => [
+            "DATE",
+            "TIME",
+            "TIMESTAMP",
+            "TIMESTAMP_MS",
+            "TIMESTAMP WITH TIME ZONE",
+        ].includes(d)).length >= 1;
+    if (doubleAndDate) {
+        throw new Error("You are trying to summarize numbers and timestamps/dates/times. You can specify values in the options (just numbers or just timestamps/dates/times) or convert your timestamps/dates/times to the number of ms since 1970-01-01 00:00:00 by passing the option { toMs: true }.");
+    }
+    const aggregates = {
+        count: "count", // specific implementation
+        countUnique: "countUnique", // specific implementation
+        countNull: "countNull", // Specific implementation
+        min: "MIN(",
+        max: "MAX(",
+        mean: "AVG(",
+        median: "MEDIAN(",
+        sum: "SUM(",
+        skew: "SKEWNESS(",
+        stdDev: "STDDEV(",
+        var: "VARIANCE(",
+    };
+    if (summaries.length === 0) {
+        summaries = Object.keys(aggregates);
+    }
+    let query = `CREATE OR REPLACE TABLE "${outputTable}" AS`;
+    let firstValue = true;
+    for (const value of values) {
+        if (firstValue) {
+            firstValue = false;
+        }
+        else {
+            query += "\nUNION";
+        }
+        query += `\nSELECT ${options.noColumnValue ? "" : `'${value}' AS 'value'`}${categories.length > 0
+            ? `, ${categories.map((d) => `"${d}"`).join(", ")}`
+            : ""},${summaries.map((summary, i) => {
+            if (value === "rowNumberToSummarizeQuerySDA" &&
+                aggregates[summary] !== "count") {
+                return `\nNULL as '${columns ? columns[i] : summary}'`;
+            }
+            else if (types[value] === "GEOMETRY") {
+                return `\nNULL AS '${columns ? columns[i] : summary}'`;
+            }
+            else if (types[value] === "VARCHAR" &&
+                [
+                    "MIN(",
+                    "MAX(",
+                    "AVG(",
+                    "MEDIAN(",
+                    "SUM(",
+                    "SKEWNESS(",
+                    "STDDEV(",
+                    "VARIANCE(",
+                ].includes(aggregates[summary])) {
+                return `\nNULL AS '${columns ? columns[i] : summary}'`;
+            }
+            else if ([
+                "DATE",
+                "TIME",
+                "TIMESTAMP",
+                "TIMESTAMP_MS",
+                "TIMESTAMP WITH TIME ZONE",
+            ].includes(types[value]) &&
+                ["AVG(", "SUM(", "SKEWNESS(", "STDDEV(", "VARIANCE("].includes(aggregates[summary])) {
+                return `\nNULL AS '${columns ? columns[i] : summary}'`;
+            }
+            else if (summary === "count") {
+                return `\nCAST(COUNT(*) AS INTEGER) AS '${columns ? columns[i] : "count"}'`;
+            }
+            else if (summary === "countUnique") {
+                return `\nCAST(COUNT(DISTINCT "${value}") AS INTEGER) AS '${columns ? columns[i] : "countUnique"}'`;
+            }
+            else if (summary === "countNull") {
+                return `\nCAST(COUNT(CASE WHEN "${value}" IS NULL THEN 1 END) AS INTEGER) as '${columns ? columns[i] : "countNull"}'`;
+            }
+            else {
+                return typeof options.decimals === "number" &&
+                    ![
+                        "VARCHAR",
+                        "DATE",
+                        "TIME",
+                        "TIMESTAMP",
+                        "TIMESTAMP WITH TIME ZONE",
+                    ].includes(types[value])
+                    ? `\nROUND(${aggregates[summary]}"${value}"), ${options.decimals}) AS '${columns ? columns[i] : summary}'`
+                    : `\n${aggregates[summary]}"${value}") AS '${columns ? columns[i] : summary}'`;
+            }
+        })}\nFROM "${table}"`;
+        if (categories.length > 0) {
+            query += `\nGROUP BY ${categories.map((d) => `"${d}"`).join(", ")}`;
+        }
+    }
+    if (options.noColumnValue) {
+        if (categories.length > 0) {
+            query += `\nORDER BY ${categories.map((d) => `"${d}" ASC`).join(", ")}`;
+        }
+    }
+    else {
+        query += `\nORDER BY ${["value", ...categories]
+            .map((d) => `"${d}" ASC`)
+            .join(", ")}`;
+    }
+    return query.replace("SELECT ,", "SELECT ");
+}

package/esm/deps/jsr.io/@nshiab/simple-data-analysis-core/0.0.17/src/methods/trimQuery.d.ts

@@ -0,0 +1,5 @@
+export default function trimQuery(table: string, columns: string[], options: {
+    character?: string;
+    method?: "leftTrim" | "rightTrim" | "trim";
+}): string;
+//# sourceMappingURL=trimQuery.d.ts.map

package/esm/deps/jsr.io/@nshiab/simple-data-analysis-core/0.0.17/src/methods/trimQuery.d.ts.map

@@ -0,0 +1 @@
+{"version":3,"file":"trimQuery.d.ts","sourceRoot":"","sources":["../../../../../../../../src/deps/jsr.io/@nshiab/simple-data-analysis-core/0.0.17/src/methods/trimQuery.ts"],"names":[],"mappings":"AAAA,MAAM,CAAC,OAAO,UAAU,SAAS,CAC/B,KAAK,EAAE,MAAM,EACb,OAAO,EAAE,MAAM,EAAE,EACjB,OAAO,EAAE;IAAE,SAAS,CAAC,EAAE,MAAM,CAAC;IAAC,MAAM,CAAC,EAAE,UAAU,GAAG,WAAW,GAAG,MAAM,CAAA;CAAE,UA8B5E"}

package/esm/deps/jsr.io/@nshiab/simple-data-analysis-core/0.0.17/src/methods/trimQuery.js

@@ -0,0 +1,29 @@
+export default function trimQuery(table, columns, options) {
+    let query = ``;
+    const method = options.method ?? "trim";
+    const specialCharacter = typeof options.character === "string"
+        ? `, '${options.character}'`
+        : "";
+    if (method === "trim") {
+        for (const column of columns) {
+            query +=
+                `\nUPDATE "${table}" SET "${column}" = TRIM("${column}"${specialCharacter});`;
+        }
+    }
+    else if (method === "leftTrim") {
+        for (const column of columns) {
+            query +=
+                `\nUPDATE "${table}" SET "${column}" = LTRIM("${column}"${specialCharacter});`;
+        }
+    }
+    else if (method === "rightTrim") {
+        for (const column of columns) {
+            query +=
+                `\nUPDATE "${table}" SET "${column}" = RTRIM("${column}"${specialCharacter});`;
+        }
+    }
+    else {
+        throw new Error(`Unknown method ${options.method}`);
+    }
+    return query;
+}

package/esm/deps/jsr.io/@nshiab/simple-data-analysis-core/0.0.17/src/methods/truncateQuery.d.ts

@@ -0,0 +1,2 @@
+export default function truncateQuery(table: string, column: string, length: number): string;
+//# sourceMappingURL=truncateQuery.d.ts.map

package/esm/deps/jsr.io/@nshiab/simple-data-analysis-core/0.0.17/src/methods/truncateQuery.d.ts.map

@@ -0,0 +1 @@
+{"version":3,"file":"truncateQuery.d.ts","sourceRoot":"","sources":["../../../../../../../../src/deps/jsr.io/@nshiab/simple-data-analysis-core/0.0.17/src/methods/truncateQuery.ts"],"names":[],"mappings":"AAAA,MAAM,CAAC,OAAO,UAAU,aAAa,CACnC,KAAK,EAAE,MAAM,EACb,MAAM,EAAE,MAAM,EACd,MAAM,EAAE,MAAM,UAGf"}

package/esm/deps/jsr.io/@nshiab/simple-data-analysis-core/0.0.17/src/methods/truncateQuery.js

@@ -0,0 +1,3 @@
+export default function truncateQuery(table, column, length) {
+    return `UPDATE "${table}" SET "${column}" = LEFT("${column}", ${length});`;
+}

package/esm/deps/jsr.io/@nshiab/simple-data-analysis-core/0.0.17/src/methods/upperQuery.d.ts

@@ -0,0 +1,2 @@
+export default function upperQuery(table: string, columns: string[]): string;
+//# sourceMappingURL=upperQuery.d.ts.map

package/esm/deps/jsr.io/@nshiab/simple-data-analysis-core/0.0.17/src/methods/upperQuery.d.ts.map

@@ -0,0 +1 @@
+{"version":3,"file":"upperQuery.d.ts","sourceRoot":"","sources":["../../../../../../../../src/deps/jsr.io/@nshiab/simple-data-analysis-core/0.0.17/src/methods/upperQuery.ts"],"names":[],"mappings":"AAAA,MAAM,CAAC,OAAO,UAAU,UAAU,CAAC,KAAK,EAAE,MAAM,EAAE,OAAO,EAAE,MAAM,EAAE,UAQlE"}

package/esm/deps/jsr.io/@nshiab/simple-data-analysis-core/0.0.17/src/methods/upperQuery.js

@@ -0,0 +1,7 @@
+export default function upperQuery(table, columns) {
+    let query = "";
+    for (const column of columns) {
+        query += `\nUPDATE "${table}" SET "${column}" = UPPER("${column}");`;
+    }
+    return query;
+}

package/esm/deps/jsr.io/@nshiab/simple-data-analysis-core/0.0.17/src/methods/writeDataQuery.d.ts

@@ -0,0 +1,5 @@
+export default function writeDataQuery(table: string, file: string, fileExtension: string, options: {
+    compression?: boolean;
+    formatDates?: boolean;
+}): string;
+//# sourceMappingURL=writeDataQuery.d.ts.map

package/esm/deps/jsr.io/@nshiab/simple-data-analysis-core/0.0.17/src/methods/writeDataQuery.d.ts.map

@@ -0,0 +1 @@
+{"version":3,"file":"writeDataQuery.d.ts","sourceRoot":"","sources":["../../../../../../../../src/deps/jsr.io/@nshiab/simple-data-analysis-core/0.0.17/src/methods/writeDataQuery.ts"],"names":[],"mappings":"AAGA,MAAM,CAAC,OAAO,UAAU,cAAc,CACpC,KAAK,EAAE,MAAM,EACb,IAAI,EAAE,MAAM,EACZ,aAAa,EAAE,MAAM,EACrB,OAAO,EAAE;IAAE,WAAW,CAAC,EAAE,OAAO,CAAC;IAAC,WAAW,CAAC,EAAE,OAAO,CAAA;CAAE,UAgD1D"}

package/esm/deps/jsr.io/@nshiab/simple-data-analysis-core/0.0.17/src/methods/writeDataQuery.js

@@ -0,0 +1,44 @@
+import { existsSync, rmSync } from "node:fs";
+import cleanPath from "../helpers/cleanPath.js";
+export default function writeDataQuery(table, file, fileExtension, options) {
+    const cleanedFile = cleanPath(file);
+    if (fileExtension === "csv") {
+        return `COPY "${table}" TO '${options.compression ? cleanedFile + ".gz" : cleanedFile}' (DELIMITER ',', HEADER TRUE${options.compression ? ", COMPRESSION GZIP" : ""}${options.formatDates
+            ? ", DATEFORMAT '%xT%X.%gZ', TIMESTAMPFORMAT '%xT%X.%gZ'"
+            : ""});`;
+    }
+    else if (fileExtension === "json") {
+        return `COPY "${table}" TO '${options.compression ? cleanedFile + ".gz" : cleanedFile}' (FORMAT JSON, ARRAY TRUE${options.compression ? ", COMPRESSION GZIP" : ""}${options.formatDates
+            ? ", DATEFORMAT '%xT%X.%gZ', TIMESTAMPFORMAT '%xT%X.%gZ'"
+            : ""});`;
+    }
+    else if (fileExtension === "parquet") {
+        if (options.compression) {
+            return `COPY "${table}" TO '${cleanedFile}' (FORMAT PARQUET, COMPRESSION ZSTD);`;
+        }
+        else {
+            return `COPY "${table}" TO '${cleanedFile}' (FORMAT PARQUET);`;
+        }
+    }
+    else if (fileExtension === "db") {
+        if (existsSync(file)) {
+            rmSync(file);
+        }
+        return `ATTACH '${cleanedFile}' AS "my_database";
+COPY FROM DATABASE memory TO "my_database";
+CREATE OR REPLACE TABLE "my_database"."${table}" AS SELECT * FROM "${table}";
+DETACH "my_database";`;
+    }
+    else if (fileExtension === "sqlite") {
+        if (existsSync(file)) {
+            rmSync(file);
+        }
+        return `INSTALL sqlite; LOAD sqlite;
+    ATTACH '${cleanedFile}' AS "my_sqlite_db" (TYPE SQLITE);
+    CREATE TABLE "my_sqlite_db"."${table}" AS SELECT * FROM "${table}";
+    DETACH "my_sqlite_db";`;
+    }
+    else {
+        throw new Error(`Unknown extension ${fileExtension}`);
+    }
+}

package/esm/deps/jsr.io/@nshiab/simple-data-analysis-core/0.0.17/src/methods/writeGeoDataQuery.d.ts

@@ -0,0 +1,4 @@
+export default function writeGeoDataQuery(table: string, file: string, fileExtension: string, options?: {
+    precision?: number;
+}): string;
+//# sourceMappingURL=writeGeoDataQuery.d.ts.map

package/esm/deps/jsr.io/@nshiab/simple-data-analysis-core/0.0.17/src/methods/writeGeoDataQuery.d.ts.map

@@ -0,0 +1 @@
+{"version":3,"file":"writeGeoDataQuery.d.ts","sourceRoot":"","sources":["../../../../../../../../src/deps/jsr.io/@nshiab/simple-data-analysis-core/0.0.17/src/methods/writeGeoDataQuery.ts"],"names":[],"mappings":"AAEA,MAAM,CAAC,OAAO,UAAU,iBAAiB,CACvC,KAAK,EAAE,MAAM,EACb,IAAI,EAAE,MAAM,EACZ,aAAa,EAAE,MAAM,EACrB,OAAO,GAAE;IAAE,SAAS,CAAC,EAAE,MAAM,CAAA;CAAO,UAqBrC"}

package/esm/deps/jsr.io/@nshiab/simple-data-analysis-core/0.0.17/src/methods/writeGeoDataQuery.js

@@ -0,0 +1,16 @@
+import cleanPath from "../helpers/cleanPath.js";
+export default function writeGeoDataQuery(table, file, fileExtension, options = {}) {
+    if (fileExtension === "geojson" || fileExtension === "json") {
+        const layerOptions = [];
+        if (typeof options.precision === "number") {
+            layerOptions.push(`COORDINATE_PRECISION=${options.precision}`);
+        }
+        layerOptions.push(`RFC7946=YES`);
+        return `COPY "${table}" to '${cleanPath(file)}' WITH (FORMAT GDAL, DRIVER 'GeoJSON'${layerOptions.length > 0
+            ? `, LAYER_CREATION_OPTIONS ('WRITE_NAME=NO', ${layerOptions.map((d) => `'${d}'`).join(", ")})`
+            : ""})`;
+    }
+    else {
+        throw new Error(`Unknown extension ${fileExtension}`);
+    }
+}

package/esm/deps/jsr.io/@nshiab/simple-data-analysis-core/0.0.17/src/methods/zScoreQuery.d.ts

@@ -0,0 +1,5 @@
+export default function zScoreQuery(table: string, column: string, newColumn: string, options?: {
+    categories?: string | string[];
+    decimals?: number;
+}): string;
+//# sourceMappingURL=zScoreQuery.d.ts.map

package/esm/deps/jsr.io/@nshiab/simple-data-analysis-core/0.0.17/src/methods/zScoreQuery.d.ts.map

@@ -0,0 +1 @@
+{"version":3,"file":"zScoreQuery.d.ts","sourceRoot":"","sources":["../../../../../../../../src/deps/jsr.io/@nshiab/simple-data-analysis-core/0.0.17/src/methods/zScoreQuery.ts"],"names":[],"mappings":"AAEA,MAAM,CAAC,OAAO,UAAU,WAAW,CACjC,KAAK,EAAE,MAAM,EACb,MAAM,EAAE,MAAM,EACd,SAAS,EAAE,MAAM,EACjB,OAAO,GAAE;IACP,UAAU,CAAC,EAAE,MAAM,GAAG,MAAM,EAAE,CAAC;IAC/B,QAAQ,CAAC,EAAE,MAAM,CAAC;CACd,UA0BP"}

package/esm/deps/jsr.io/@nshiab/simple-data-analysis-core/0.0.17/src/methods/zScoreQuery.js

@@ -0,0 +1,23 @@
+import stringToArray from "../helpers/stringToArray.js";
+export default function zScoreQuery(table, column, newColumn, options = {}) {
+    const categories = options.categories
+        ? stringToArray(options.categories)
+        : [];
+    const partition = categories.length > 0
+        ? `PARTITION BY ${categories.map((d) => `"${d}"`).join(", ")}`
+        : "";
+    const tempQuery = `("${column}"-AVG("${column}") OVER(${partition}))
+            /
+            STDDEV_POP("${column}") OVER(${partition})`;
+    const query = `
+    CREATE OR REPLACE TABLE "${table}" AS
+    SELECT *, (
+        ${typeof options.decimals === "number"
+        ? `ROUND(${tempQuery}, ${options.decimals})`
+        : tempQuery}
+        
+        ) AS ${newColumn},
+    FROM "${table}"
+    `;
+    return query;
+}

package/esm/helpers/getRRFRanking.d.ts

@@ -0,0 +1,8 @@
+/**
+ * Executes Reciprocal Rank Fusion and returns only the sorted document IDs.
+ * @param results - An array of string arrays (e.g., [[id1, id2], [id2, id3]])
+ * @param k - The smoothing constant (default 60)
+ * @returns A flat array of document IDs sorted by their fused rank.
+ */
+export default function getRRFRanking(results: string[][], k?: number): string[];
+//# sourceMappingURL=getRRFRanking.d.ts.map

package/esm/helpers/getRRFRanking.d.ts.map

@@ -0,0 +1 @@
+{"version":3,"file":"getRRFRanking.d.ts","sourceRoot":"","sources":["../../src/helpers/getRRFRanking.ts"],"names":[],"mappings":"AAAA;;;;;GAKG;AACH,MAAM,CAAC,OAAO,UAAU,aAAa,CACnC,OAAO,EAAE,MAAM,EAAE,EAAE,EACnB,CAAC,GAAE,MAAW,GACb,MAAM,EAAE,CAiBV"}

package/esm/helpers/getRRFRanking.js

@@ -0,0 +1,21 @@
+/**
+ * Executes Reciprocal Rank Fusion and returns only the sorted document IDs.
+ * @param results - An array of string arrays (e.g., [[id1, id2], [id2, id3]])
+ * @param k - The smoothing constant (default 60)
+ * @returns A flat array of document IDs sorted by their fused rank.
+ */
+export default function getRRFRanking(results, k = 60) {
+    const scoreMap = new Map();
+    results.forEach((resultSet) => {
+        resultSet.forEach((id, index) => {
+            const rank = index + 1;
+            const currentScore = scoreMap.get(id) || 0;
+            // Add the RRF score to the existing total for this ID
+            scoreMap.set(id, currentScore + (1 / (k + rank)));
+        });
+    });
+    // Sort by score descending and return only the IDs
+    return Array.from(scoreMap.entries())
+        .sort((a, b) => b[1] - a[1])
+        .map(([id]) => id);
+}

package/esm/helpers/sleep.d.ts

@@ -0,0 +1,5 @@
+export default function sleep(ms: number, options?: {
+    start?: Date;
+    log?: boolean;
+}): Promise<void>;
+//# sourceMappingURL=sleep.d.ts.map

package/esm/helpers/sleep.d.ts.map

@@ -0,0 +1 @@
+{"version":3,"file":"sleep.d.ts","sourceRoot":"","sources":["../../src/helpers/sleep.ts"],"names":[],"mappings":"AAEA,MAAM,CAAC,OAAO,UAAU,KAAK,CAC3B,EAAE,EAAE,MAAM,EACV,OAAO,GAAE;IAAE,KAAK,CAAC,EAAE,IAAI,CAAC;IAAC,GAAG,CAAC,EAAE,OAAO,CAAA;CAAO,GAC5C,OAAO,CAAC,IAAI,CAAC,CA0Bf"}

package/esm/helpers/sleep.js

@@ -0,0 +1,23 @@
+import { prettyDuration } from "../deps/jsr.io/@nshiab/journalism-format/1.1.5/src/index.js";
+export default function sleep(ms, options = {}) {
+    return new Promise((resolve) => {
+        if (options.start) {
+            const end = new Date();
+            const duration = end.getTime() - options.start.getTime();
+            const remaining = ms - duration;
+            if (remaining > 0) {
+                if (options.log) {
+                    console.log(`\nSleeping for ${prettyDuration(0, { end: remaining })}...`);
+                }
+                return setTimeout(resolve, remaining);
+            }
+            else {
+                if (options.log) {
+                    console.log(`\nNo need to sleep, already took ${prettyDuration(0, { end: duration })}.`);
+                }
+                return resolve();
+            }
+        }
+        return setTimeout(resolve, ms);
+    });
+}

package/esm/helpers/tryAI.d.ts

@@ -0,0 +1,36 @@
+import type { Ollama } from "ollama";
+export default function tryAI(i: number, batchSize: number, rows: {
+    [key: string]: string | number | boolean | Date | null;
+}[], column: string, newColumns: string[], prompt: string, options?: {
+    batchSize?: number;
+    concurrent?: number;
+    cache?: boolean;
+    test?: (result: {
+        [key: string]: unknown;
+    }) => void;
+    retry?: number;
+    model?: string;
+    temperature?: number;
+    apiKey?: string;
+    vertex?: boolean;
+    project?: string;
+    location?: string;
+    ollama?: boolean | Ollama;
+    verbose?: boolean;
+    rateLimitPerMinute?: number;
+    clean?: (response: unknown) => unknown;
+    contextWindow?: number;
+    thinkingBudget?: number;
+    thinkingLevel?: "minimal" | "low" | "medium" | "high";
+    safetyEnabled?: boolean;
+    webSearch?: boolean;
+    extraInstructions?: string;
+    schemaJson?: unknown;
+    metrics?: {
+        totalCost: number;
+        totalInputTokens: number;
+        totalOutputTokens: number;
+        totalRequests: number;
+    };
+}): Promise<void>;
+//# sourceMappingURL=tryAI.d.ts.map

package/esm/helpers/tryAI.d.ts.map

@@ -0,0 +1 @@
+{"version":3,"file":"tryAI.d.ts","sourceRoot":"","sources":["../../src/helpers/tryAI.ts"],"names":[],"mappings":"AACA,OAAO,KAAK,EAAE,MAAM,EAAE,MAAM,QAAQ,CAAC;AAGrC,wBAA8B,KAAK,CACjC,CAAC,EAAE,MAAM,EACT,SAAS,EAAE,MAAM,EACjB,IAAI,EAAE;IACJ,CAAC,GAAG,EAAE,MAAM,GAAG,MAAM,GAAG,MAAM,GAAG,OAAO,GAAG,IAAI,GAAG,IAAI,CAAC;CACxD,EAAE,EACH,MAAM,EAAE,MAAM,EACd,UAAU,EAAE,MAAM,EAAE,EACpB,MAAM,EAAE,MAAM,EACd,OAAO,GAAE;IACP,SAAS,CAAC,EAAE,MAAM,CAAC;IACnB,UAAU,CAAC,EAAE,MAAM,CAAC;IACpB,KAAK,CAAC,EAAE,OAAO,CAAC;IAChB,IAAI,CAAC,EAAE,CAAC,MAAM,EAAE;QAAE,CAAC,GAAG,EAAE,MAAM,GAAG,OAAO,CAAA;KAAE,KAAK,IAAI,CAAC;IACpD,KAAK,CAAC,EAAE,MAAM,CAAC;IACf,KAAK,CAAC,EAAE,MAAM,CAAC;IACf,WAAW,CAAC,EAAE,MAAM,CAAC;IACrB,MAAM,CAAC,EAAE,MAAM,CAAC;IAChB,MAAM,CAAC,EAAE,OAAO,CAAC;IACjB,OAAO,CAAC,EAAE,MAAM,CAAC;IACjB,QAAQ,CAAC,EAAE,MAAM,CAAC;IAClB,MAAM,CAAC,EAAE,OAAO,GAAG,MAAM,CAAC;IAC1B,OAAO,CAAC,EAAE,OAAO,CAAC;IAClB,kBAAkB,CAAC,EAAE,MAAM,CAAC;IAC5B,KAAK,CAAC,EAAE,CACN,QAAQ,EAAE,OAAO,KACd,OAAO,CAAC;IACb,aAAa,CAAC,EAAE,MAAM,CAAC;IACvB,cAAc,CAAC,EAAE,MAAM,CAAC;IACxB,aAAa,CAAC,EAAE,SAAS,GAAG,KAAK,GAAG,QAAQ,GAAG,MAAM,CAAC;IACtD,aAAa,CAAC,EAAE,OAAO,CAAC;IACxB,SAAS,CAAC,EAAE,OAAO,CAAC;IACpB,iBAAiB,CAAC,EAAE,MAAM,CAAC;IAC3B,UAAU,CAAC,EAAE,OAAO,CAAC;IACrB,OAAO,CAAC,EAAE;QACR,SAAS,EAAE,MAAM,CAAC;QAClB,gBAAgB,EAAE,MAAM,CAAC;QACzB,iBAAiB,EAAE,MAAM,CAAC;QAC1B,aAAa,EAAE,MAAM,CAAC;KACvB,CAAC;CACE,iBAgHP"}

package/esm/helpers/tryAI.js

@@ -0,0 +1,75 @@
+import { askAI } from "../deps/jsr.io/@nshiab/journalism-ai/1.2.18/src/index.js";
+import { array, object, string, toJSONSchema } from "zod";
+export default async function tryAI(i, batchSize, rows, column, newColumns, prompt, options = {}) {
+    const batch = rows.slice(i, i + batchSize);
+    let schemaJson;
+    if (options.schemaJson) {
+        schemaJson = options.schemaJson;
+    }
+    else {
+        const objectSchema = {};
+        for (const newColumn of newColumns) {
+            objectSchema[newColumn] = string();
+        }
+        schemaJson = toJSONSchema(array(object(objectSchema)));
+    }
+    const systemPrompt = `You will be provided with a JSON array of ${batch.length} string items. You must return a JSON array containing exactly ${batch.length} objects, in the same corresponding order.`;
+    const fullPrompt = `${prompt}\n\nHere are the ${column} values as a JSON array:\n${JSON.stringify(batch.map((d) => d[column]), null, 2)}\n\n${options.extraInstructions ? `\n${options.extraInstructions}` : ""}`;
+    const retry = options.retry ?? 1;
+    let testPassed = false;
+    let iterations = 1;
+    let newValues;
+    while (!testPassed && iterations <= retry) {
+        try {
+            newValues = await askAI(`${fullPrompt}${iterations > 1
+                ? `\nThis is your attempt #${iterations}. So get it right by following my instructions closely!`
+                : ""}`, {
+                ...options,
+                includeThoughts: options.verbose ? true : false,
+                systemPrompt: systemPrompt,
+                schemaJson,
+                test: (response) => {
+                    if (!Array.isArray(response)) {
+                        throw new Error(`The AI returned a non-array value: ${JSON.stringify(response)}`);
+                    }
+                    if (response.length !== batch.length) {
+                        throw new Error(`The AI returned ${response.length} values, but the batch size is ${batch.length}.`);
+                    }
+                    for (const item of response) {
+                        if (typeof item !== "object" || item === null) {
+                            throw new Error(`The AI did not return an object: ${JSON.stringify(item)}`);
+                        }
+                        for (const newColumn of newColumns) {
+                            if (!(newColumn in item)) {
+                                throw new Error(`The AI's response is missing the key '${newColumn}': ${JSON.stringify(item)}`);
+                            }
+                        }
+                    }
+                    if (options.test) {
+                        for (const item of response) {
+                            options.test(item);
+                        }
+                    }
+                },
+            });
+            testPassed = true;
+        }
+        catch (e) {
+            if (iterations < retry) {
+                console.log(`${console.log(e)}\nRetrying... (${iterations}/${retry})`);
+                iterations++;
+            }
+            else {
+                console.log(`${console.log(e)}\nNo more retries left. (${iterations}/${retry}).`);
+                throw e;
+            }
+        }
+    }
+    // Types could be improved
+    for (let j = 0; j <
+        newValues.length; j++) {
+        for (const newColumn of newColumns) {
+            rows[i + j][newColumn] = newValues[j][newColumn];
+        }
+    }
+}

package/esm/helpers/tryEmbedding.d.ts

@@ -0,0 +1,15 @@
+import type { Ollama } from "ollama";
+export default function tryEmbedding(i: number, rows: {
+    [key: string]: string | number | boolean | Date | null;
+}[], text: string, newColumn: string, options?: {
+    cache?: boolean;
+    model?: string;
+    apiKey?: string;
+    vertex?: boolean;
+    project?: string;
+    location?: string;
+    ollama?: boolean | Ollama;
+    verbose?: boolean;
+    contextWindow?: number;
+}): Promise<number>;
+//# sourceMappingURL=tryEmbedding.d.ts.map

package/esm/helpers/tryEmbedding.d.ts.map

@@ -0,0 +1 @@
+{"version":3,"file":"tryEmbedding.d.ts","sourceRoot":"","sources":["../../src/helpers/tryEmbedding.ts"],"names":[],"mappings":"AACA,OAAO,KAAK,EAAE,MAAM,EAAE,MAAM,QAAQ,CAAC;AAErC,wBAA8B,YAAY,CACxC,CAAC,EAAE,MAAM,EACT,IAAI,EAAE;IACJ,CAAC,GAAG,EAAE,MAAM,GAAG,MAAM,GAAG,MAAM,GAAG,OAAO,GAAG,IAAI,GAAG,IAAI,CAAC;CACxD,EAAE,EACH,IAAI,EAAE,MAAM,EACZ,SAAS,EAAE,MAAM,EACjB,OAAO,GAAE;IACP,KAAK,CAAC,EAAE,OAAO,CAAC;IAChB,KAAK,CAAC,EAAE,MAAM,CAAC;IACf,MAAM,CAAC,EAAE,MAAM,CAAC;IAChB,MAAM,CAAC,EAAE,OAAO,CAAC;IACjB,OAAO,CAAC,EAAE,MAAM,CAAC;IACjB,QAAQ,CAAC,EAAE,MAAM,CAAC;IAClB,MAAM,CAAC,EAAE,OAAO,GAAG,MAAM,CAAC;IAC1B,OAAO,CAAC,EAAE,OAAO,CAAC;IAClB,aAAa,CAAC,EAAE,MAAM,CAAC;CACnB,mBAOP"}

package/esm/helpers/tryEmbedding.js

@@ -0,0 +1,5 @@
+import { getEmbedding } from "../deps/jsr.io/@nshiab/journalism-ai/1.2.18/src/index.js";
+export default async function tryEmbedding(i, rows, text, newColumn, options = {}) {
+    // Should be improved...
+    return rows[i][newColumn] = await getEmbedding(text, options);
+}

package/esm/index.d.ts

@@ -0,0 +1,31 @@
+/**
+ * @module
+ *
+ * The Simple Data Analysis Library
+ *
+ * To install the library with Deno, use:
+ * ```bash
+ * deno add jsr:@nshiab/simple-data-analysis
+ * ```
+ *
+ * To install the library with Node.js, use:
+ * ```bash
+ * npx jsr add @nshiab/simple-data-analysis
+ * ```
+ *
+ * To start, create a SimpleDB instance and then a SimpleTable from this instance:
+ * ```ts
+ * import { SimpleDB } from "@nshiab/simple-data-analysis";
+ *
+ * const sdb = new SimpleDB();
+ * const table = sdb.newTable("myTable"); // This returns a SimpleTable instance
+ * await table.loadData("path/to/your/data.csv");
+ *
+ * // You can now perform various data analysis operations on the table.
+ *
+ * await sdb.done(); // Ensure to call done when you're finished.
+ * ```
+ */
+export { default as SimpleDB } from "./class/SimpleDB.js";
+export { default as SimpleTable } from "./class/SimpleTable.js";
+//# sourceMappingURL=index.d.ts.map

package/esm/index.d.ts.map

@@ -0,0 +1 @@
+{"version":3,"file":"index.d.ts","sourceRoot":"","sources":["../src/index.ts"],"names":[],"mappings":"AAAA;;;;;;;;;;;;;;;;;;;;;;;;;;;GA2BG;AAEH,OAAO,EAAE,OAAO,IAAI,QAAQ,EAAE,MAAM,qBAAqB,CAAC;AAC1D,OAAO,EAAE,OAAO,IAAI,WAAW,EAAE,MAAM,wBAAwB,CAAC"}

package/esm/index.js

@@ -0,0 +1,30 @@
+/**
+ * @module
+ *
+ * The Simple Data Analysis Library
+ *
+ * To install the library with Deno, use:
+ * ```bash
+ * deno add jsr:@nshiab/simple-data-analysis
+ * ```
+ *
+ * To install the library with Node.js, use:
+ * ```bash
+ * npx jsr add @nshiab/simple-data-analysis
+ * ```
+ *
+ * To start, create a SimpleDB instance and then a SimpleTable from this instance:
+ * ```ts
+ * import { SimpleDB } from "@nshiab/simple-data-analysis";
+ *
+ * const sdb = new SimpleDB();
+ * const table = sdb.newTable("myTable"); // This returns a SimpleTable instance
+ * await table.loadData("path/to/your/data.csv");
+ *
+ * // You can now perform various data analysis operations on the table.
+ *
+ * await sdb.done(); // Ensure to call done when you're finished.
+ * ```
+ */
+export { default as SimpleDB } from "./class/SimpleDB.js";
+export { default as SimpleTable } from "./class/SimpleTable.js";

package/esm/methods/aiEmbeddings.d.ts

@@ -0,0 +1,21 @@
+import type { SimpleTable } from "../index.js";
+import type { Ollama } from "ollama";
+export default function aiEmbeddings(simpleTable: SimpleTable, column: string, newColumn: string, options?: {
+    createIndex?: boolean;
+    overwriteIndex?: boolean;
+    concurrent?: number;
+    cache?: boolean;
+    model?: string;
+    apiKey?: string;
+    vertex?: boolean;
+    project?: string;
+    location?: string;
+    ollama?: boolean | Ollama;
+    verbose?: boolean;
+    rateLimitPerMinute?: number;
+    contextWindow?: number;
+    efConstruction?: number;
+    efSearch?: number;
+    M?: number;
+}): Promise<void>;
+//# sourceMappingURL=aiEmbeddings.d.ts.map

package/esm/methods/aiEmbeddings.d.ts.map

@@ -0,0 +1 @@
+{"version":3,"file":"aiEmbeddings.d.ts","sourceRoot":"","sources":["../../src/methods/aiEmbeddings.ts"],"names":[],"mappings":"AAEA,OAAO,KAAK,EAAE,WAAW,EAAE,MAAM,aAAa,CAAC;AAE/C,OAAO,KAAK,EAAE,MAAM,EAAE,MAAM,QAAQ,CAAC;AAErC,wBAA8B,YAAY,CACxC,WAAW,EAAE,WAAW,EACxB,MAAM,EAAE,MAAM,EACd,SAAS,EAAE,MAAM,EACjB,OAAO,GAAE;IACP,WAAW,CAAC,EAAE,OAAO,CAAC;IACtB,cAAc,CAAC,EAAE,OAAO,CAAC;IACzB,UAAU,CAAC,EAAE,MAAM,CAAC;IACpB,KAAK,CAAC,EAAE,OAAO,CAAC;IAChB,KAAK,CAAC,EAAE,MAAM,CAAC;IACf,MAAM,CAAC,EAAE,MAAM,CAAC;IAChB,MAAM,CAAC,EAAE,OAAO,CAAC;IACjB,OAAO,CAAC,EAAE,MAAM,CAAC;IACjB,QAAQ,CAAC,EAAE,MAAM,CAAC;IAClB,MAAM,CAAC,EAAE,OAAO,GAAG,MAAM,CAAC;IAC1B,OAAO,CAAC,EAAE,OAAO,CAAC;IAClB,kBAAkB,CAAC,EAAE,MAAM,CAAC;IAC5B,aAAa,CAAC,EAAE,MAAM,CAAC;IACvB,cAAc,CAAC,EAAE,MAAM,CAAC;IACxB,QAAQ,CAAC,EAAE,MAAM,CAAC;IAClB,CAAC,CAAC,EAAE,MAAM,CAAC;CACP,iBAsEP"}